drm/amdgpu: add amdgpu_ras.c to support ras (v2)

add obj management.
add feature control.
add debugfs infrastructure.
add sysfs infrastructure.
add IH infrastructure.
add recovery infrastructure.

It is a framework. Other IPs need call amdgpu_ras_xxx function instead of
psp_ras_xxx functions.

v2: squash in warning fixes
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/amd/amdgpu/Makefile

@@ -53,7 +53,7 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
 	amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \
 	amdgpu_gtt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o amdgpu_atomfirmware.o \
 	amdgpu_vf_error.o amdgpu_sched.o amdgpu_debugfs.o amdgpu_ids.o \
-	amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o
+	amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o amdgpu_ras.o
 
 # add asic specific block
 amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -60,6 +60,7 @@
 #include "amdgpu_pm.h"
 
 #include "amdgpu_xgmi.h"
+#include "amdgpu_ras.h"
 
 MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/vega12_gpu_info.bin");
@@ -1638,6 +1639,10 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 {
 	int i, r;
 
+	r = amdgpu_ras_init(adev);
+	if (r)
+		return r;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@@ -1876,6 +1881,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 {
 	int i, r;
 
+	amdgpu_ras_pre_fini(adev);
+
 	if (adev->gmc.xgmi.num_physical_nodes > 1)
 		amdgpu_xgmi_remove_device(adev);
 
@@ -1945,6 +1952,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 		adev->ip_blocks[i].status.late_initialized = false;
 	}
 
+	amdgpu_ras_fini(adev);
+
 	if (amdgpu_sriov_vf(adev))
 		if (amdgpu_virt_release_full_gpu(adev, false))
 			DRM_ERROR("failed to release exclusive mode on fini\n");

drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h

@@ -110,6 +110,7 @@ struct psp_ras_context {
 	struct amdgpu_bo	*ras_shared_bo;
 	uint64_t		ras_shared_mc_addr;
 	void			*ras_shared_buf;
+	struct amdgpu_ras	*ras;
 };
 
 struct psp_context

drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h

+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ *
+ */
+#ifndef _AMDGPU_RAS_H
+#define _AMDGPU_RAS_H
+
+#include <linux/debugfs.h>
+#include <linux/list.h>
+#include "amdgpu.h"
+#include "amdgpu_psp.h"
+#include "ta_ras_if.h"
+
+enum amdgpu_ras_block {
+	AMDGPU_RAS_BLOCK__UMC = 0,
+	AMDGPU_RAS_BLOCK__SDMA,
+	AMDGPU_RAS_BLOCK__GFX,
+	AMDGPU_RAS_BLOCK__MMHUB,
+	AMDGPU_RAS_BLOCK__ATHUB,
+	AMDGPU_RAS_BLOCK__PCIE_BIF,
+	AMDGPU_RAS_BLOCK__HDP,
+	AMDGPU_RAS_BLOCK__XGMI_WAFL,
+	AMDGPU_RAS_BLOCK__DF,
+	AMDGPU_RAS_BLOCK__SMN,
+	AMDGPU_RAS_BLOCK__SEM,
+	AMDGPU_RAS_BLOCK__MP0,
+	AMDGPU_RAS_BLOCK__MP1,
+	AMDGPU_RAS_BLOCK__FUSE,
+
+	AMDGPU_RAS_BLOCK__LAST
+};
+
+#define AMDGPU_RAS_BLOCK_COUNT	AMDGPU_RAS_BLOCK__LAST
+#define AMDGPU_RAS_BLOCK_MASK	((1ULL << AMDGPU_RAS_BLOCK_COUNT) - 1)
+
+enum amdgpu_ras_error_type {
+	AMDGPU_RAS_ERROR__NONE							= 0,
+	AMDGPU_RAS_ERROR__PARITY						= 1,
+	AMDGPU_RAS_ERROR__SINGLE_CORRECTABLE					= 2,
+	AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE					= 4,
+	AMDGPU_RAS_ERROR__POISON						= 8,
+};
+
+enum amdgpu_ras_ret {
+	AMDGPU_RAS_SUCCESS = 0,
+	AMDGPU_RAS_FAIL,
+	AMDGPU_RAS_UE,
+	AMDGPU_RAS_CE,
+	AMDGPU_RAS_PT,
+};
+
+struct ras_common_if {
+	enum amdgpu_ras_block block;
+	enum amdgpu_ras_error_type type;
+	uint32_t sub_block_index;
+	/* block name */
+	char name[32];
+};
+
+typedef int (*ras_ih_cb)(struct amdgpu_device *adev,
+		struct amdgpu_iv_entry *entry);
+
+struct amdgpu_ras {
+	/* ras infrastructure */
+	uint32_t supported;
+	uint32_t features;
+	struct list_head head;
+	/* debugfs */
+	struct dentry *dir;
+	/* sysfs */
+	struct device_attribute features_attr;
+	/* block array */
+	struct ras_manager *objs;
+
+	/* gpu recovery */
+	struct work_struct recovery_work;
+	atomic_t in_recovery;
+	struct amdgpu_device *adev;
+	/* error handler data */
+	struct ras_err_handler_data *eh_data;
+	struct mutex recovery_lock;
+};
+
+/* interfaces for IP */
+
+struct ras_fs_if {
+	struct ras_common_if head;
+	char sysfs_name[32];
+	char debugfs_name[32];
+};
+
+struct ras_query_if {
+	struct ras_common_if head;
+	unsigned long ue_count;
+	unsigned long ce_count;
+};
+
+struct ras_inject_if {
+	struct ras_common_if head;
+	uint64_t address;
+	uint64_t value;
+};
+
+struct ras_cure_if {
+	struct ras_common_if head;
+	uint64_t address;
+};
+
+struct ras_ih_if {
+	struct ras_common_if head;
+	ras_ih_cb cb;
+};
+
+struct ras_dispatch_if {
+	struct ras_common_if head;
+	struct amdgpu_iv_entry *entry;
+};
+
+/* work flow
+ * vbios
+ * 1: ras feature enable (enabled by default)
+ * psp
+ * 2: ras framework init (in ip_init)
+ * IP
+ * 3: IH add
+ * 4: debugfs/sysfs create
+ * 5: query/inject
+ * 6: debugfs/sysfs remove
+ * 7: IH remove
+ * 8: feature disable
+ */
+
+#define amdgpu_ras_get_context(adev)		((adev)->psp.ras.ras)
+#define amdgpu_ras_set_context(adev, ras_con)	((adev)->psp.ras.ras = (ras_con))
+
+/* check if ras is supported on block, say, sdma, gfx */
+static inline int amdgpu_ras_is_supported(struct amdgpu_device *adev,
+		unsigned int block)
+{
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	return ras && (ras->supported & (1 << block));
+}
+
+int amdgpu_ras_query_error_count(struct amdgpu_device *adev,
+		bool is_ce);
+
+/* error handling functions */
+int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev,
+		unsigned long *bps, int pages);
+
+int amdgpu_ras_reserve_bad_pages(struct amdgpu_device *adev);
+
+static inline int amdgpu_ras_reset_gpu(struct amdgpu_device *adev,
+		bool is_baco)
+{
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	if (atomic_cmpxchg(&ras->in_recovery, 0, 1) == 0)
+		schedule_work(&ras->recovery_work);
+	return 0;
+}
+
+/* called in ip_init and ip_fini */
+int amdgpu_ras_init(struct amdgpu_device *adev);
+int amdgpu_ras_fini(struct amdgpu_device *adev);
+int amdgpu_ras_pre_fini(struct amdgpu_device *adev);
+
+int amdgpu_ras_feature_enable(struct amdgpu_device *adev,
+		struct ras_common_if *head, bool enable);
+
+int amdgpu_ras_sysfs_create(struct amdgpu_device *adev,
+		struct ras_fs_if *head);
+
+int amdgpu_ras_sysfs_remove(struct amdgpu_device *adev,
+		struct ras_common_if *head);
+
+int amdgpu_ras_debugfs_create(struct amdgpu_device *adev,
+		struct ras_fs_if *head);
+
+int amdgpu_ras_debugfs_remove(struct amdgpu_device *adev,
+		struct ras_common_if *head);
+
+int amdgpu_ras_error_query(struct amdgpu_device *adev,
+		struct ras_query_if *info);
+
+int amdgpu_ras_error_inject(struct amdgpu_device *adev,
+		struct ras_inject_if *info);
+
+int amdgpu_ras_interrupt_add_handler(struct amdgpu_device *adev,
+		struct ras_ih_if *info);
+
+int amdgpu_ras_interrupt_remove_handler(struct amdgpu_device *adev,
+		struct ras_ih_if *info);
+
+int amdgpu_ras_interrupt_dispatch(struct amdgpu_device *adev,
+		struct ras_dispatch_if *info);
+#endif