summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915/i915_gpu_error.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/i915/i915_gpu_error.c')
-rw-r--r--drivers/gpu/drm/i915/i915_gpu_error.c73
1 files changed, 24 insertions, 49 deletions
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 4eb58887819a..0e4b832dff84 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -726,13 +726,6 @@ static void err_print_gt_info(struct drm_i915_error_state_buf *m,
intel_sseu_print_topology(gt->_gt->i915, &gt->info.sseu, &p);
}
-static void err_print_gt_display(struct drm_i915_error_state_buf *m,
- struct intel_gt_coredump *gt)
-{
- err_printf(m, "IER: 0x%08x\n", gt->ier);
- err_printf(m, "DERRMR: 0x%08x\n", gt->derrmr);
-}
-
static void err_print_gt_global_nonguc(struct drm_i915_error_state_buf *m,
struct intel_gt_coredump *gt)
{
@@ -841,7 +834,6 @@ static void __err_print_to_sgl(struct drm_i915_error_state_buf *m,
err_printf(m, "Kernel: %s %s\n",
init_utsname()->release,
init_utsname()->machine);
- err_printf(m, "Driver: %s\n", DRIVER_DATE);
ts = ktime_to_timespec64(error->time);
err_printf(m, "Time: %lld s %ld us\n",
(s64)ts.tv_sec, ts.tv_nsec / NSEC_PER_USEC);
@@ -879,7 +871,6 @@ static void __err_print_to_sgl(struct drm_i915_error_state_buf *m,
if (error->gt->uc && error->gt->uc->guc.is_guc_capture)
print_guc_capture = true;
- err_print_gt_display(m, error->gt);
err_print_gt_global_nonguc(m, error->gt);
err_print_gt_fences(m, error->gt);
@@ -1643,9 +1634,21 @@ capture_engine(struct intel_engine_cs *engine,
return NULL;
intel_engine_get_hung_entity(engine, &ce, &rq);
- if (rq && !i915_request_started(rq))
- drm_info(&engine->gt->i915->drm, "Got hung context on %s with active request %lld:%lld [0x%04X] not yet started\n",
- engine->name, rq->fence.context, rq->fence.seqno, ce->guc_id.id);
+ if (rq && !i915_request_started(rq)) {
+ /*
+ * We want to know also what is the guc_id of the context,
+ * but if we don't have the context reference, then skip
+ * printing it.
+ */
+ if (ce)
+ drm_info(&engine->gt->i915->drm,
+ "Got hung context on %s with active request %lld:%lld [0x%04X] not yet started\n",
+ engine->name, rq->fence.context, rq->fence.seqno, ce->guc_id.id);
+ else
+ drm_info(&engine->gt->i915->drm,
+ "Got hung context on %s with active request %lld:%lld not yet started\n",
+ engine->name, rq->fence.context, rq->fence.seqno);
+ }
if (rq) {
capture = intel_engine_coredump_add_request(ee, rq, ATOMIC_MAYFAIL);
@@ -1756,27 +1759,6 @@ gt_record_uc(struct intel_gt_coredump *gt,
return error_uc;
}
-/* Capture display registers. */
-static void gt_record_display_regs(struct intel_gt_coredump *gt)
-{
- struct intel_uncore *uncore = gt->_gt->uncore;
- struct drm_i915_private *i915 = uncore->i915;
-
- if (DISPLAY_VER(i915) >= 6 && DISPLAY_VER(i915) < 20)
- gt->derrmr = intel_uncore_read(uncore, DERRMR);
-
- if (GRAPHICS_VER(i915) >= 8)
- gt->ier = intel_uncore_read(uncore, GEN8_DE_MISC_IER);
- else if (IS_VALLEYVIEW(i915))
- gt->ier = intel_uncore_read(uncore, VLV_IER);
- else if (HAS_PCH_SPLIT(i915))
- gt->ier = intel_uncore_read(uncore, DEIER);
- else if (GRAPHICS_VER(i915) == 2)
- gt->ier = intel_uncore_read16(uncore, GEN2_IER);
- else
- gt->ier = intel_uncore_read(uncore, GEN2_IER);
-}
-
/* Capture all other registers that GuC doesn't capture. */
static void gt_record_global_nonguc_regs(struct intel_gt_coredump *gt)
{
@@ -1810,9 +1792,12 @@ static void gt_record_global_nonguc_regs(struct intel_gt_coredump *gt)
gt->gtier[i] =
intel_uncore_read(uncore, GEN8_GT_IER(i));
gt->ngtier = 4;
- } else if (HAS_PCH_SPLIT(i915)) {
+ } else if (GRAPHICS_VER(i915) >= 5) {
gt->gtier[0] = intel_uncore_read(uncore, GTIER);
gt->ngtier = 1;
+ } else {
+ gt->gtier[0] = intel_uncore_read(uncore, GEN2_IER);
+ gt->ngtier = 1;
}
gt->eir = intel_uncore_read(uncore, EIR);
@@ -2032,7 +2017,6 @@ intel_gt_coredump_alloc(struct intel_gt *gt, gfp_t gfp, u32 dump_flags)
gc->_gt = gt;
gc->awake = intel_gt_pm_is_awake(gt);
- gt_record_display_regs(gc);
gt_record_global_nonguc_regs(gc);
/*
@@ -2083,7 +2067,7 @@ static struct i915_gpu_coredump *
__i915_gpu_coredump(struct intel_gt *gt, intel_engine_mask_t engine_mask, u32 dump_flags)
{
struct drm_i915_private *i915 = gt->i915;
- struct intel_display *display = &i915->display;
+ struct intel_display *display = i915->display;
struct i915_gpu_coredump *error;
/* Check if GPU capture has been disabled */
@@ -2149,7 +2133,6 @@ i915_gpu_coredump(struct intel_gt *gt, intel_engine_mask_t engine_mask, u32 dump
void i915_error_state_store(struct i915_gpu_coredump *error)
{
struct drm_i915_private *i915;
- static bool warned;
if (IS_ERR_OR_NULL(error))
return;
@@ -2163,16 +2146,8 @@ void i915_error_state_store(struct i915_gpu_coredump *error)
i915_gpu_coredump_get(error);
- if (!xchg(&warned, true) &&
- ktime_get_real_seconds() - DRIVER_TIMESTAMP < DAY_AS_SECONDS(180)) {
- pr_info("GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.\n");
- pr_info("Please file a _new_ bug report at https://gitlab.freedesktop.org/drm/intel/issues/new.\n");
- pr_info("Please see https://drm.pages.freedesktop.org/intel-docs/how-to-file-i915-bugs.html for details.\n");
- pr_info("drm/i915 developers can then reassign to the right component if it's not a kernel issue.\n");
- pr_info("The GPU crash dump is required to analyze GPU hangs, so please always attach it.\n");
- pr_info("GPU crash dump saved to /sys/class/drm/card%d/error\n",
- i915->drm.primary->index);
- }
+ drm_info(&i915->drm, "GPU error state saved to /sys/class/drm/card%d/error\n",
+ i915->drm.primary->index);
}
/**
@@ -2479,7 +2454,7 @@ void i915_gpu_error_debugfs_register(struct drm_i915_private *i915)
}
static ssize_t error_state_read(struct file *filp, struct kobject *kobj,
- struct bin_attribute *attr, char *buf,
+ const struct bin_attribute *attr, char *buf,
loff_t off, size_t count)
{
@@ -2515,7 +2490,7 @@ static ssize_t error_state_read(struct file *filp, struct kobject *kobj,
}
static ssize_t error_state_write(struct file *file, struct kobject *kobj,
- struct bin_attribute *attr, char *buf,
+ const struct bin_attribute *attr, char *buf,
loff_t off, size_t count)
{
struct device *kdev = kobj_to_dev(kobj);
generated by cgit v1.2.3 (git 2.39.0) at 2025-08-11 15:24:16 +0300