| Message ID | 20240109034755.100555-2-Smita.KoralahalliChannabasappa@amd.com |
|---|---|
| State | New, archived |
| Headers | show |
| Series | acpi/ghes, cper, cxl: Trace FW-First CXL Protocol Errors | expand |
On Tue, 9 Jan 2024 03:47:52 +0000 Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com> wrote: > Currently defined cxl_cper_callback interface between CXL subsystem and > GHES module is just confined to handling CXL Component errors only. > > Extend this callback to process CXL Protocol errors as well. Achieve > by defining a new struct cxl_cper_event_info to include cxl_cper_event_rec > and other fields of CXL protocol errors which will be defined in future > patches. > > Signed-off-by: Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com> Hi Smita, I guess this will get effected by the mess around the reporting that Ira is fixing but in meantime some comments on the current code. > --- > v2: > cxl_cper_rec_data -> cxl_cper_event_info > data -> info > --- > drivers/acpi/apei/ghes.c | 6 +++++- > drivers/cxl/pci.c | 8 ++++---- > include/linux/cxl-event.h | 6 +++++- > 3 files changed, 14 insertions(+), 6 deletions(-) > > diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c > index aed465d2fd68..60b615d361d3 100644 > --- a/drivers/acpi/apei/ghes.c > +++ b/drivers/acpi/apei/ghes.c > @@ -693,6 +693,10 @@ static cxl_cper_callback cper_callback; > static void cxl_cper_post_event(enum cxl_event_type event_type, > struct cxl_cper_event_rec *rec) > { > + struct cxl_cper_event_info info; > + > + info.rec = *(struct cxl_cper_event_rec *)rec; Why cast? > + > if (rec->hdr.length <= sizeof(rec->hdr) || > rec->hdr.length > sizeof(*rec)) { > pr_err(FW_WARN "CXL CPER Invalid section length (%u)\n", > @@ -707,7 +711,7 @@ static void cxl_cper_post_event(enum cxl_event_type event_type, > > guard(rwsem_read)(&cxl_cper_rw_sem); > if (cper_callback) > - cper_callback(event_type, rec); > + cper_callback(event_type, &info); > } > > int cxl_cper_register_callback(cxl_cper_callback callback) > diff --git a/drivers/cxl/pci.c b/drivers/cxl/pci.c > index b14237f824cf..1ad240ead4fd 100644 > --- a/drivers/cxl/pci.c > +++ b/drivers/cxl/pci.c > @@ -972,9 +972,9 @@ static struct pci_driver cxl_pci_driver = { > > #define CXL_EVENT_HDR_FLAGS_REC_SEVERITY GENMASK(1, 0) > static void cxl_cper_event_call(enum cxl_event_type ev_type, > - struct cxl_cper_event_rec *rec) > + struct cxl_cper_event_info *info) > { > - struct cper_cxl_event_devid *device_id = &rec->hdr.device_id; > + struct cper_cxl_event_devid *device_id = &info->rec.hdr.device_id; > struct pci_dev *pdev __free(pci_dev_put) = NULL; > enum cxl_event_log_type log_type; > struct cxl_dev_state *cxlds; > @@ -996,11 +996,11 @@ static void cxl_cper_event_call(enum cxl_event_type ev_type, > return; > > /* Fabricate a log type */ > - hdr_flags = get_unaligned_le24(rec->event.generic.hdr.flags); > + hdr_flags = get_unaligned_le24(info->rec.event.generic.hdr.flags); > log_type = FIELD_GET(CXL_EVENT_HDR_FLAGS_REC_SEVERITY, hdr_flags); > > cxl_event_trace_record(cxlds->cxlmd, log_type, ev_type, > - &uuid_null, &rec->event); > + &uuid_null, &info->rec.event); > } > > static int __init cxl_pci_driver_init(void) > diff --git a/include/linux/cxl-event.h b/include/linux/cxl-event.h > index 17eadee819b6..6ce839c59749 100644 > --- a/include/linux/cxl-event.h > +++ b/include/linux/cxl-event.h > @@ -141,8 +141,12 @@ struct cxl_cper_event_rec { > union cxl_event event; > } __packed; > > +struct cxl_cper_event_info { > + struct cxl_cper_event_rec rec; Only parts of this will be relevant to the protocol errors. Maybe worth doing a union with the first part of rec in both structures but not the union cxl_event in the protocol error. Keep it all anonymous to avoid yet another structure in the reads/and writes though. > +}; > + > typedef void (*cxl_cper_callback)(enum cxl_event_type type, > - struct cxl_cper_event_rec *rec); > + struct cxl_cper_event_info *info); > > #ifdef CONFIG_ACPI_APEI_GHES > int cxl_cper_register_callback(cxl_cper_callback callback);
diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c index aed465d2fd68..60b615d361d3 100644 --- a/drivers/acpi/apei/ghes.c +++ b/drivers/acpi/apei/ghes.c @@ -693,6 +693,10 @@ static cxl_cper_callback cper_callback; static void cxl_cper_post_event(enum cxl_event_type event_type, struct cxl_cper_event_rec *rec) { + struct cxl_cper_event_info info; + + info.rec = *(struct cxl_cper_event_rec *)rec; + if (rec->hdr.length <= sizeof(rec->hdr) || rec->hdr.length > sizeof(*rec)) { pr_err(FW_WARN "CXL CPER Invalid section length (%u)\n", @@ -707,7 +711,7 @@ static void cxl_cper_post_event(enum cxl_event_type event_type, guard(rwsem_read)(&cxl_cper_rw_sem); if (cper_callback) - cper_callback(event_type, rec); + cper_callback(event_type, &info); } int cxl_cper_register_callback(cxl_cper_callback callback) diff --git a/drivers/cxl/pci.c b/drivers/cxl/pci.c index b14237f824cf..1ad240ead4fd 100644 --- a/drivers/cxl/pci.c +++ b/drivers/cxl/pci.c @@ -972,9 +972,9 @@ static struct pci_driver cxl_pci_driver = { #define CXL_EVENT_HDR_FLAGS_REC_SEVERITY GENMASK(1, 0) static void cxl_cper_event_call(enum cxl_event_type ev_type, - struct cxl_cper_event_rec *rec) + struct cxl_cper_event_info *info) { - struct cper_cxl_event_devid *device_id = &rec->hdr.device_id; + struct cper_cxl_event_devid *device_id = &info->rec.hdr.device_id; struct pci_dev *pdev __free(pci_dev_put) = NULL; enum cxl_event_log_type log_type; struct cxl_dev_state *cxlds; @@ -996,11 +996,11 @@ static void cxl_cper_event_call(enum cxl_event_type ev_type, return; /* Fabricate a log type */ - hdr_flags = get_unaligned_le24(rec->event.generic.hdr.flags); + hdr_flags = get_unaligned_le24(info->rec.event.generic.hdr.flags); log_type = FIELD_GET(CXL_EVENT_HDR_FLAGS_REC_SEVERITY, hdr_flags); cxl_event_trace_record(cxlds->cxlmd, log_type, ev_type, - &uuid_null, &rec->event); + &uuid_null, &info->rec.event); } static int __init cxl_pci_driver_init(void) diff --git a/include/linux/cxl-event.h b/include/linux/cxl-event.h index 17eadee819b6..6ce839c59749 100644 --- a/include/linux/cxl-event.h +++ b/include/linux/cxl-event.h @@ -141,8 +141,12 @@ struct cxl_cper_event_rec { union cxl_event event; } __packed; +struct cxl_cper_event_info { + struct cxl_cper_event_rec rec; +}; + typedef void (*cxl_cper_callback)(enum cxl_event_type type, - struct cxl_cper_event_rec *rec); + struct cxl_cper_event_info *info); #ifdef CONFIG_ACPI_APEI_GHES int cxl_cper_register_callback(cxl_cper_callback callback);
Currently defined cxl_cper_callback interface between CXL subsystem and GHES module is just confined to handling CXL Component errors only. Extend this callback to process CXL Protocol errors as well. Achieve by defining a new struct cxl_cper_event_info to include cxl_cper_event_rec and other fields of CXL protocol errors which will be defined in future patches. Signed-off-by: Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com> --- v2: cxl_cper_rec_data -> cxl_cper_event_info data -> info --- drivers/acpi/apei/ghes.c | 6 +++++- drivers/cxl/pci.c | 8 ++++---- include/linux/cxl-event.h | 6 +++++- 3 files changed, 14 insertions(+), 6 deletions(-)