Commit 365aa9f5 authored by Dave Airlie's avatar Dave Airlie
Browse files

Merge tag 'amd-drm-next-6.11-2024-06-22' of... 

Merge tag 'amd-drm-next-6.11-2024-06-22' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.11-2024-06-22:

amdgpu:
- HPD fixes
- PSR fixes
- DCC updates
- DCN 4.0.1 fixes
- FAMS fixes
- Misc code cleanups
- SR-IOV fixes
- GPUVM TLB flush cleanups
- Make VCN less verbose
- ACPI backlight fixes
- MES fixes
- Firmware loading cleanups
- Replay fixes
- LTTPR fixes
- Trap handler fixes
- Cursor and overlay fixes
- Primary plane zpos fixes
- DML 2.1 fixes
- RAS updates
- USB4 fixes
- MALL fixes
- Reserved VMID fix
- Silence UBSAN warnings

amdkfd:
- Misc code cleanups

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240622152523.2267072-1-alexander.deucher@amd.com


Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
parents 541b1b0a 1ecef558

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

+2 −0
Original line number Diff line number Diff line
@@ -1441,6 +1441,7 @@ u32 amdgpu_device_pcie_port_rreg(struct amdgpu_device *adev, 
				u32 reg);

void amdgpu_device_pcie_port_wreg(struct amdgpu_device *adev,

				u32 reg, u32 v);

struct dma_fence *amdgpu_device_get_gang(struct amdgpu_device *adev);

struct dma_fence *amdgpu_device_switch_gang(struct amdgpu_device *adev,

					    struct dma_fence *gang);

bool amdgpu_device_has_display_hardware(struct amdgpu_device *adev);
@@ -1567,6 +1568,7 @@ static inline int amdgpu_acpi_power_shift_control(struct amdgpu_device *adev, 
						  u8 dev_state, bool drv_state) { return 0; }

static inline int amdgpu_acpi_smart_shift_update(struct drm_device *dev,

						 enum amdgpu_ss ss_state) { return 0; }

static inline void amdgpu_acpi_get_backlight_caps(struct amdgpu_dm_backlight_caps *caps) { }

#endif



#if defined(CONFIG_ACPI) && defined(CONFIG_SUSPEND)

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

+20 −12
Original line number Diff line number Diff line
@@ -222,9 +222,9 @@ static struct aca_bank_error *new_bank_error(struct aca_error *aerr, struct aca_ 
	INIT_LIST_HEAD(&bank_error->node);

	memcpy(&bank_error->info, info, sizeof(*info));



	spin_lock(&aerr->lock);

	mutex_lock(&aerr->lock);

	list_add_tail(&bank_error->node, &aerr->list);

	spin_unlock(&aerr->lock);

	mutex_unlock(&aerr->lock);



	return bank_error;

}
@@ -235,7 +235,7 @@ static struct aca_bank_error *find_bank_error(struct aca_error *aerr, struct aca 
	struct aca_bank_info *tmp_info;

	bool found = false;



	spin_lock(&aerr->lock);

	mutex_lock(&aerr->lock);

	list_for_each_entry(bank_error, &aerr->list, node) {

		tmp_info = &bank_error->info;

		if (tmp_info->socket_id == info->socket_id &&
@@ -246,7 +246,7 @@ static struct aca_bank_error *find_bank_error(struct aca_error *aerr, struct aca 
	}



out_unlock:

	spin_unlock(&aerr->lock);

	mutex_unlock(&aerr->lock);



	return found ? bank_error : NULL;

}
@@ -474,7 +474,7 @@ static int aca_log_aca_error(struct aca_handle *handle, enum aca_error_type type 
	struct aca_error *aerr = &error_cache->errors[type];

	struct aca_bank_error *bank_error, *tmp;



	spin_lock(&aerr->lock);

	mutex_lock(&aerr->lock);



	if (list_empty(&aerr->list))

		goto out_unlock;
@@ -485,7 +485,7 @@ static int aca_log_aca_error(struct aca_handle *handle, enum aca_error_type type 
	}



out_unlock:

	spin_unlock(&aerr->lock);

	mutex_unlock(&aerr->lock);



	return 0;

}
@@ -542,7 +542,7 @@ int amdgpu_aca_get_error_data(struct amdgpu_device *adev, struct aca_handle *han 


static void aca_error_init(struct aca_error *aerr, enum aca_error_type type)

{

	spin_lock_init(&aerr->lock);

	mutex_init(&aerr->lock);

	INIT_LIST_HEAD(&aerr->list);

	aerr->type = type;

	aerr->nr_errors = 0;
@@ -561,10 +561,11 @@ static void aca_error_fini(struct aca_error *aerr) 
{

	struct aca_bank_error *bank_error, *tmp;



	spin_lock(&aerr->lock);

	mutex_lock(&aerr->lock);

	list_for_each_entry_safe(bank_error, tmp, &aerr->list, node)

		aca_bank_error_remove(aerr, bank_error);

	spin_unlock(&aerr->lock);



	mutex_destroy(&aerr->lock);

}



static void aca_fini_error_cache(struct aca_handle *handle)
@@ -712,6 +713,15 @@ void amdgpu_aca_fini(struct amdgpu_device *adev) 
	atomic_set(&aca->ue_update_flag, 0);

}



int amdgpu_aca_reset(struct amdgpu_device *adev)

{

	struct amdgpu_aca *aca = &adev->aca;



	atomic_set(&aca->ue_update_flag, 0);



	return 0;

}



void amdgpu_aca_set_smu_funcs(struct amdgpu_device *adev, const struct aca_smu_funcs *smu_funcs)

{

	struct amdgpu_aca *aca = &adev->aca;
@@ -885,9 +895,7 @@ DEFINE_DEBUGFS_ATTRIBUTE(aca_debug_mode_fops, NULL, amdgpu_aca_smu_debug_mode_se 
void amdgpu_aca_smu_debugfs_init(struct amdgpu_device *adev, struct dentry *root)

{

#if defined(CONFIG_DEBUG_FS)

	if (!root ||

	    (adev->ip_versions[MP1_HWIP][0] != IP_VERSION(13, 0, 6) &&

	     adev->ip_versions[MP1_HWIP][0] != IP_VERSION(13, 0, 14)))

	if (!root)

		return;



	debugfs_create_file("aca_debug_mode", 0200, root, adev, &aca_debug_mode_fops);

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

+2 −2
Original line number Diff line number Diff line
@@ -25,7 +25,6 @@ 
#define __AMDGPU_ACA_H__



#include <linux/list.h>

#include <linux/spinlock.h>



struct ras_err_data;

struct ras_query_context;
@@ -134,7 +133,7 @@ struct aca_bank_error { 


struct aca_error {

	struct list_head list;

	spinlock_t lock;

	struct mutex lock;

	enum aca_error_type type;

	int nr_errors;

};
@@ -192,6 +191,7 @@ struct aca_info { 


int amdgpu_aca_init(struct amdgpu_device *adev);

void amdgpu_aca_fini(struct amdgpu_device *adev);

int amdgpu_aca_reset(struct amdgpu_device *adev);

void amdgpu_aca_set_smu_funcs(struct amdgpu_device *adev, const struct aca_smu_funcs *smu_funcs);

bool amdgpu_aca_is_enabled(struct amdgpu_device *adev);

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

+4 −0
Original line number Diff line number Diff line
@@ -383,6 +383,8 @@ static int amdgpu_atif_query_backlight_caps(struct amdgpu_atif *atif) 
			characteristics.min_input_signal;

	atif->backlight_caps.max_input_signal =

			characteristics.max_input_signal;

	atif->backlight_caps.ac_level = characteristics.ac_level;

	atif->backlight_caps.dc_level = characteristics.dc_level;

out:

	kfree(info);

	return err;
@@ -1268,6 +1270,8 @@ void amdgpu_acpi_get_backlight_caps(struct amdgpu_dm_backlight_caps *caps) 
	caps->caps_valid = atif->backlight_caps.caps_valid;

	caps->min_input_signal = atif->backlight_caps.min_input_signal;

	caps->max_input_signal = atif->backlight_caps.max_input_signal;

	caps->ac_level = atif->backlight_caps.ac_level;

	caps->dc_level = atif->backlight_caps.dc_level;

}



/**

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

+25 −20
Original line number Diff line number Diff line
@@ -2350,7 +2350,6 @@ void amdgpu_device_set_sriov_virtual_display(struct amdgpu_device *adev) 
static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)

{

	const char *chip_name;

	char fw_name[40];

	int err;

	const struct gpu_info_firmware_header_v1_0 *hdr;
@@ -2384,12 +2383,12 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev) 
		break;

	}



	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_gpu_info.bin", chip_name);

	err = amdgpu_ucode_request(adev, &adev->firmware.gpu_info_fw, fw_name);

	err = amdgpu_ucode_request(adev, &adev->firmware.gpu_info_fw,

				   "amdgpu/%s_gpu_info.bin", chip_name);

	if (err) {

		dev_err(adev->dev,

			"Failed to get gpu_info firmware \"%s\"\n",

			fw_name);

			"Failed to get gpu_info firmware \"%s_gpu_info.bin\"\n",

			chip_name);

		goto out;

	}
@@ -5070,6 +5069,8 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, 
	struct amdgpu_hive_info *hive = NULL;



	if (test_bit(AMDGPU_HOST_FLR, &reset_context->flags)) {

		amdgpu_virt_ready_to_reset(adev);

		amdgpu_virt_wait_reset(adev);

		clear_bit(AMDGPU_HOST_FLR, &reset_context->flags);

		r = amdgpu_virt_request_full_gpu(adev, true);

	} else {
@@ -6275,20 +6276,11 @@ pci_ers_result_t amdgpu_pci_slot_reset(struct pci_dev *pdev) 
	struct amdgpu_reset_context reset_context;

	u32 memsize;

	struct list_head device_list;

	struct amdgpu_hive_info *hive;

	int hive_ras_recovery = 0;

	struct amdgpu_ras *ras;



	/* PCI error slot reset should be skipped During RAS recovery */

	hive = amdgpu_get_xgmi_hive(adev);

	if (hive) {

		hive_ras_recovery = atomic_read(&hive->ras_recovery);

		amdgpu_put_xgmi_hive(hive);

	}

	ras = amdgpu_ras_get_context(adev);

	if ((amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 3) ||

	    amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 4)) &&

	    ras && (atomic_read(&ras->in_recovery) || hive_ras_recovery))

	    amdgpu_ras_in_recovery(adev))

		return PCI_ERS_RESULT_RECOVERED;



	DRM_INFO("PCI error: slot reset callback!!\n");
@@ -6530,6 +6522,22 @@ void amdgpu_device_pcie_port_wreg(struct amdgpu_device *adev, 
	spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);

}



/**

 * amdgpu_device_get_gang - return a reference to the current gang

 * @adev: amdgpu_device pointer

 *

 * Returns: A new reference to the current gang leader.

 */

struct dma_fence *amdgpu_device_get_gang(struct amdgpu_device *adev)

{

	struct dma_fence *fence;



	rcu_read_lock();

	fence = dma_fence_get_rcu_safe(&adev->gang_submit);

	rcu_read_unlock();

	return fence;

}



/**

 * amdgpu_device_switch_gang - switch to a new gang

 * @adev: amdgpu_device pointer
@@ -6546,10 +6554,7 @@ struct dma_fence *amdgpu_device_switch_gang(struct amdgpu_device *adev, 


	do {

		dma_fence_put(old);

		rcu_read_lock();

		old = dma_fence_get_rcu_safe(&adev->gang_submit);

		rcu_read_unlock();



		old = amdgpu_device_get_gang(adev);

		if (old == gang)

			break;