Read CPER raw data from debugfs node "/sys/kernel/debug/dri/*/ amdgpu_ring_cper".
Signed-off-by: Xiang Liu <[email protected]> Reviewed-by: Tao Zhou <[email protected]> Reviewed-by: Yang Wang <[email protected]> --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + .../gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c | 3 +- drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c | 87 +++++++++++++++++++ 3 files changed, 90 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 603cc155fd9b..66ccd31f1e8e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -110,6 +110,7 @@ #include "amdgpu_mca.h" #include "amdgpu_aca.h" #include "amdgpu_ras.h" +#include "amdgpu_ras_mgr.h" #include "amdgpu_cper.h" #include "amdgpu_xcp.h" #include "amdgpu_seq64.h" diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c index 5a7bf0661dbf..011fa4748084 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c @@ -770,7 +770,8 @@ amdgpu_ras_eeprom_update_header(struct amdgpu_ras_eeprom_control *control) "Saved bad pages %d reaches threshold value %d\n", control->ras_num_bad_pages, ras->bad_page_cnt_threshold); - if (adev->cper.enabled && amdgpu_cper_generate_bp_threshold_record(adev)) + if (adev->cper.enabled && !amdgpu_uniras_enabled(adev) && + amdgpu_cper_generate_bp_threshold_record(adev)) dev_warn(adev->dev, "fail to generate bad page threshold cper records\n"); if ((amdgpu_bad_page_threshold != -1) && diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index f26002b25de4..8637d924cd48 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -491,6 +491,90 @@ bool amdgpu_ring_soft_recovery(struct amdgpu_ring *ring, unsigned int vmid, */ #if defined(CONFIG_DEBUG_FS) +static ssize_t amdgpu_ras_cper_debugfs_read(struct file *f, char __user *buf, + size_t size, loff_t *offset) +{ + struct amdgpu_ring *ring = file_inode(f)->i_private; + struct ras_cmd_cper_snapshot_req *snapshot_req; + struct ras_cmd_cper_snapshot_rsp *snapshot_rsp; + struct ras_cmd_cper_record_req *record_req; + struct ras_cmd_cper_record_rsp *record_rsp; + const uint8_t ring_header_size = 12; + uint8_t *ring_header; + uint32_t total_cper_num; + uint64_t start_cper_id; + int r; + + ring_header = kzalloc(ring_header_size, GFP_KERNEL); + if (!ring_header) { + r = -ENOMEM; + goto out; + } + + snapshot_req = kzalloc(sizeof(struct ras_cmd_cper_snapshot_req), GFP_KERNEL); + if (!snapshot_req) { + r = -ENOMEM; + goto out; + } + + snapshot_rsp = kzalloc(sizeof(struct ras_cmd_cper_snapshot_rsp), GFP_KERNEL); + if (!snapshot_rsp) { + r = -ENOMEM; + goto out; + } + + record_req = kzalloc(sizeof(struct ras_cmd_cper_record_req), GFP_KERNEL); + if (!record_req) { + r = -ENOMEM; + goto out; + } + + record_rsp = kzalloc(sizeof(struct ras_cmd_cper_record_rsp), GFP_KERNEL); + if (!record_rsp) { + r = -ENOMEM; + goto out; + } + + if (!(*offset)) { + if (copy_to_user(buf, ring_header, ring_header_size)) { + r = -EFAULT; + goto out; + } + buf += ring_header_size; + } + + r = amdgpu_ras_mgr_handle_ras_cmd(ring->adev, + RAS_CMD__GET_CPER_SNAPSHOT, + snapshot_req, sizeof(struct ras_cmd_cper_snapshot_req), + snapshot_rsp, sizeof(struct ras_cmd_cper_snapshot_rsp)); + if (r || !snapshot_rsp->total_cper_num) + goto out; + + start_cper_id = snapshot_rsp->start_cper_id; + total_cper_num = snapshot_rsp->total_cper_num; + + record_req->buf_ptr = (uint64_t)(uintptr_t)buf; + record_req->buf_size = size; + record_req->cper_start_id = start_cper_id + *offset; + record_req->cper_num = total_cper_num; + r = amdgpu_ras_mgr_handle_ras_cmd(ring->adev, RAS_CMD__GET_CPER_RECORD, + record_req, sizeof(struct ras_cmd_cper_record_req), + record_rsp, sizeof(struct ras_cmd_cper_record_rsp)); + if (r) + goto out; + + r = *offset ? record_rsp->real_data_size : record_rsp->real_data_size + ring_header_size; + (*offset) += record_rsp->real_cper_num; + +out: + kfree(record_rsp); + kfree(record_req); + kfree(snapshot_rsp); + kfree(snapshot_req); + kfree(ring_header); + return r; +} + /* Layout of file is 12 bytes consisting of * - rptr * - wptr @@ -507,6 +591,9 @@ static ssize_t amdgpu_debugfs_ring_read(struct file *f, char __user *buf, loff_t i; int r; + if (ring->funcs->type == AMDGPU_RING_TYPE_CPER && amdgpu_uniras_enabled(ring->adev)) + return amdgpu_ras_cper_debugfs_read(f, buf, size, pos); + if (*pos & 3 || size & 3) return -EINVAL; -- 2.34.1
