Re: [PATCH v2 03/11] drm/xe/cper: Add CPER structures and trace event

[Rodrigo Vivi <rodrigo.vivi@intel.com>]

On Tue, Aug 25, 2026 at 11:29:20PM +0530, Badal Nilawar wrote:
> Define packed data structures and Intel-specific GUID macros needed
> to build Intel GPU CPER (Common Platform Error Record) non-standard
> records.
> 
> Add xe_error_cper trace event to log the assembled CPER record bytes.
> 
> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com>
> Assisted-by: Copilot:claude-sonnet-4.6
> ---
>  drivers/gpu/drm/xe/Makefile        |   1 +
>  drivers/gpu/drm/xe/xe_cper_types.h | 186 +++++++++++++++++++++++++++++
>  drivers/gpu/drm/xe/xe_trace_cper.c |   9 ++
>  drivers/gpu/drm/xe/xe_trace_cper.h |  66 ++++++++++
>  4 files changed, 262 insertions(+)
>  create mode 100644 drivers/gpu/drm/xe/xe_cper_types.h
>  create mode 100644 drivers/gpu/drm/xe/xe_trace_cper.c
>  create mode 100644 drivers/gpu/drm/xe/xe_trace_cper.h
> 
> diff --git a/drivers/gpu/drm/xe/Makefile b/drivers/gpu/drm/xe/Makefile
> index 92134709d998..3ed60697f3f3 100644
> --- a/drivers/gpu/drm/xe/Makefile
> +++ b/drivers/gpu/drm/xe/Makefile
> @@ -136,6 +136,7 @@ xe-y += xe_bb.o \
>  	xe_tlb_inval_job.o \
>  	xe_trace.o \
>  	xe_trace_bo.o \
> +	xe_trace_cper.o \
>  	xe_trace_guc.o \
>  	xe_trace_lrc.o \
>  	xe_ttm_stolen_mgr.o \
> diff --git a/drivers/gpu/drm/xe/xe_cper_types.h b/drivers/gpu/drm/xe/xe_cper_types.h
> new file mode 100644
> index 000000000000..82167ea4eb16
> --- /dev/null
> +++ b/drivers/gpu/drm/xe/xe_cper_types.h
> @@ -0,0 +1,186 @@
> +/* SPDX-License-Identifier: MIT */
> +/*
> + * Copyright © 2026 Intel Corporation
> + */
> +
> +#ifndef _XE_CPER_TYPES_H_
> +#define _XE_CPER_TYPES_H_
> +
> +#include <linux/cper.h>
> +#include <linux/types.h>
> +#include <linux/uuid.h>
> +
> +/* Intel CPER GUID Namespace — RFC 9562 UUIDv5 (SHA-1 name-based)
> + *
> + * All values below are generated deterministically by the shell script
> + * in the [Generation Script] section. Re-run that script to verify.
> + * Do NOT hand-edit the byte values.
> + */
> +
> +/* Creator IDs */
> +#define INTEL_CPER_CREATOR_XEKMD \
> +	GUID_INIT(0x9a42070f, 0xdf9d, 0x555e, \
> +		  0xba, 0x02, 0x7c, 0xbc, 0x86, 0x3d, 0x37, 0x1c)
> +
> +#define INTEL_CPER_CREATOR_AMC \
> +	GUID_INIT(0x215803da, 0xfc7a, 0x5925, \
> +		  0xb7, 0x8b, 0x1f, 0xc1, 0x19, 0x61, 0x58, 0xd1)
> +
> +/* Notification Types */
> +#define INTEL_CPER_NOTIFY_GPU_ERROR \
> +	GUID_INIT(0x4ae12aef, 0x8745, 0x5fc7, \
> +		  0xb9, 0x96, 0x71, 0xee, 0xbb, 0x51, 0xf2, 0x23)
> +
> +#define INTEL_CPER_NOTIFY_DRV_ERROR \
> +	GUID_INIT(0xcef7e934, 0x51e7, 0x535f, \
> +		  0xa6, 0x78, 0x5a, 0x4c, 0xcc, 0xb6, 0x96, 0x09)
> +
> +/* Section Types */
> +#define INTEL_CPER_SECTION_ACCEL_GENERIC \
> +	GUID_INIT(0xea9d8f84, 0x4258, 0x5227, \
> +		  0x80, 0x28, 0xb9, 0xb1, 0x3e, 0x6d, 0x58, 0xb0)

Where all these numbers come from?

> +
> +#pragma pack(push, 1)
> +
> +/**
> + * struct xe_cper_sec_intel_err_hdr - Intel-specific CPER error section header
> + *
> + * Fixed-size header for the Intel GPU error section of a CPER record.
> + * All multi-byte fields are little-endian; the structure is packed.
> + */
> +struct xe_cper_sec_intel_err_hdr {
> +	/** @error_class: Error classification (type, component, location, cause) */
> +	union {
> +		struct {
> +			/** @error_class.error_type: RAS error severity */
> +			u8  error_type;
> +			/** @error_class.error_component: IP block that raised the error */
> +			u8  error_component;
> +			/** @error_class.tile: Tile number */
> +			u8  tile;
> +			/** @error_class.instance: Instance within the tile */
> +			u32 instance;
> +			/** @error_class.cause: Error cause code */
> +			u32 cause;
> +			/** @error_class.reserved: Reserved, must be zero */
> +			u8  reserved;
> +		} error_class;
> +		/** @class: Raw byte view of the error class */
> +		u8 class[12];
> +	};
> +	/** @first_timestamp: Timestamp of the first occurrence of this error class */
> +	u64 first_timestamp;
> +	/** @sig_id: Aggregated error class SIG ID; set to U32_MAX if unknown */
> +	u32 sig_id;
> +	/** @error_count: Number of times this error has been observed */
> +	u32 error_count;
> +	/** @valid_bits: Bitmask indicating which header fields are populated */
> +	union {
> +		struct {
> +			/** @valid_bits.location: @error_class field is valid */
> +			u16 location		: 1;
> +			/** @valid_bits.first_timestamp: @first_timestamp field is valid */
> +			u16 first_timestamp	: 1;
> +			/** @valid_bits.sig_id: @sig_id field is valid */
> +			u16 sig_id			: 1;
> +			/** @valid_bits.pci_bdf: @pci_bdf field is valid */
> +			u16 pci_bdf		: 1;
> +			/** @valid_bits.drv_version: @drv_version field is valid */
> +			u16 drv_version		: 1;
> +			/** @valid_bits.fw_id: @fw_id field is valid */
> +			u16 fw_id			: 1;
> +			/** @valid_bits.reserved: Reserved, must be zero */
> +			u16 reserved		: 10;
> +		} valid_bits;
> +		/** @validation_bits: Raw u16 view of all valid bits */
> +		u16 validation_bits;
> +	};
> +	/** @pci_bdf: PCI location string, format "DDDD:bb:dd.f" */
> +	char pci_bdf[16];
> +	/** @drv_version: Driver source version string (THIS_MODULE->srcversion) */
> +	char drv_version[25];
> +	/** @fw_id: Firmware version string (GFSP+PCODE+CSC+GUC or MNG+NUC+RAS+GUC) */
> +	char fw_id[256];
> +	/** @reserved: Reserved for future use, must be zero */
> +	u8 reserved[5];
> +};
> +
> +/**
> + * struct xe_cper_sec_intel_error_info - Variable-length Intel GPU error payload
> + *
> + * Appended after &xe_cper_sec_intel_err_hdr when detailed per-event data
> + * is available. The @event_queue flexible array holds @event_queue_count
> + * packed &xe_intel_priv_event_entry records.
> + */
> +struct xe_cper_sec_intel_error_info {
> +	/** @error_class: Error classification (mirrors the header error_class) */
> +	union {
> +		struct {
> +			u8  error_type;
> +			u8  error_component;
> +			u8  tile;
> +			u32 instance;
> +			u32 cause;
> +			u8  reserved;
> +		} error_class;
> +		/** @class: Raw byte view of the error class */
> +		u8 class[12];
> +	};
> +	/** @error_count: Total number of errors recorded */
> +	u32 error_count;
> +	/** @event_queue_length: Total byte size of the @event_queue array */
> +	u32 event_queue_length;
> +	/** @event_queue_count: Number of entries in @event_queue */
> +	u32 event_queue_count;
> +	/** @event_queue: Packed array of &xe_intel_priv_event_entry records */
> +	u8 event_queue[];
> +};
> +
> +/**
> + * struct xe_intel_priv_event_entry - Single error event in the event queue
> + *
> + * Each entry is variable-length; @entry_length gives the byte size of
> + * @metadata only (not including @entry_length or @timestamp).
> + */
> +struct xe_intel_priv_event_entry {
> +	/** @entry_length: Byte length of the @metadata payload */
> +	u32 entry_length;
> +	/** @timestamp: Hardware timestamp of this event */
> +	u64 timestamp;
> +	/** @metadata: Event-specific payload bytes */
> +	u8  metadata[];
> +};
> +
> +/**
> + * struct xe_cper_nonstd_record - Fixed-size portion of an Intel GPU CPER record
> + *
> + * Contains the standard CPER record header, section descriptor, and the
> + * Intel error section header. A &xe_cper_sec_intel_error_info payload
> + * (with its flexible @event_queue array) is appended dynamically.
> + */
> +struct xe_cper_nonstd_record {
> +	/** @record_hdr: Standard CPER record header (UEFI Appendix N.2.1) */
> +	struct cper_record_header record_hdr;
> +	/** @section_desc: CPER section descriptor */
> +	struct cper_section_descriptor section_desc;
> +	/** @intel_hdr: Intel-specific error section header */
> +	struct xe_cper_sec_intel_err_hdr intel_hdr;
> +};
> +
> +#pragma pack(pop)
> +
> +/**
> + * struct xe_platform_id_entry - Mapping from PCI device ID to CPER platform GUID
> + *
> + * Used to resolve the platform_id field in a CPER section descriptor.
> + * GUIDs are UUIDv5 (RFC 9562, SHA-1) derived from the Intel CPER namespace
> + * with name string "platform/8086:<dev_id_hex_lower>".
> + */
> +struct xe_platform_id_entry {
> +	/** @device_id: PCI device ID */
> +	u16       device_id;
> +	/** @platform_id: Corresponding UUIDv5 platform GUID */
> +	guid_t    platform_id;
> +};
> +
> +#endif
> diff --git a/drivers/gpu/drm/xe/xe_trace_cper.c b/drivers/gpu/drm/xe/xe_trace_cper.c
> new file mode 100644
> index 000000000000..caea8783ab7c
> --- /dev/null
> +++ b/drivers/gpu/drm/xe/xe_trace_cper.c
> @@ -0,0 +1,9 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/*
> + * Copyright © 2026 Intel Corporation
> + */
> +
> +#ifndef __CHECKER__
> +#define CREATE_TRACE_POINTS
> +#include "xe_trace_cper.h"
> +#endif
> diff --git a/drivers/gpu/drm/xe/xe_trace_cper.h b/drivers/gpu/drm/xe/xe_trace_cper.h
> new file mode 100644
> index 000000000000..6d2dbf504888
> --- /dev/null
> +++ b/drivers/gpu/drm/xe/xe_trace_cper.h
> @@ -0,0 +1,66 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright © 2026 Intel Corporation
> + */
> +
> +#undef TRACE_SYSTEM
> +#define TRACE_SYSTEM xe
> +
> +#if !defined(_XE_TRACE_CPER_H_) || defined(TRACE_HEADER_MULTI_READ)
> +#define _XE_TRACE_CPER_H_
> +
> +#include <linux/tracepoint.h>
> +#include <linux/types.h>
> +
> +#include "xe_cper_types.h"
> +#include "xe_device_types.h"
> +
> +#define __dev_name_xe(xe)	dev_name((xe)->drm.dev)
> +
> +TRACE_EVENT(xe_error_cper,
> +	TP_PROTO(struct xe_device *xe,
> +	    const guid_t *platform_id, const guid_t *fru_id,
> +	    const u8 severity,
> +	    const struct xe_cper_sec_intel_err_hdr *ihdr,
> +	    u32 cper_len, const u8 *cper),
> +	TP_ARGS(xe, platform_id, fru_id, severity, ihdr, cper_len, cper),
> +
> +	TP_STRUCT__entry(
> +	    __string(dev, __dev_name_xe(xe))
> +	    __array(char, platform_id, UUID_SIZE)
> +	    __array(char, fru_id, UUID_SIZE)
> +	    __field(u8, sev)
> +	    __array(u8, ihdr_raw, sizeof(struct xe_cper_sec_intel_err_hdr))
> +	    __field(u32, cper_len)
> +	    __dynamic_array(u8, cper, cper_len)
> +	    ),
> +
> +	TP_fast_assign(
> +	    __assign_str(dev);
> +	    __entry->sev = severity;
> +	    memcpy(__entry->platform_id, platform_id, UUID_SIZE);
> +	    memcpy(__entry->fru_id, fru_id, UUID_SIZE);
> +	    memcpy(__entry->ihdr_raw, ihdr, sizeof(struct xe_cper_sec_intel_err_hdr));
> +	    __entry->cper_len = cper_len;
> +	    memcpy(__get_dynamic_array(cper), cper, cper_len);
> +	    ),
> +
> +	TP_printk("dev=%s severity=%d platform_id=%pU fru_id=%pU "
> +		"intel_err_hdr_raw=%s cper_len=%u cper_raw=%s",
> +		__get_str(dev), __entry->sev,
> +		__entry->platform_id, __entry->fru_id,
> +		__print_hex(__entry->ihdr_raw,
> +		    sizeof(struct xe_cper_sec_intel_err_hdr)),
> +		__entry->cper_len,
> +		__print_hex(__get_dynamic_array(cper),
> +		    __entry->cper_len))
> +);
> +
> +#endif
> +
> +/* This part must be outside protection */
> +#undef TRACE_INCLUDE_PATH
> +#undef TRACE_INCLUDE_FILE
> +#define TRACE_INCLUDE_PATH ../../drivers/gpu/drm/xe
> +#define TRACE_INCLUDE_FILE xe_trace_cper
> +#include <trace/define_trace.h>
> -- 
> 2.54.0
>