diff options
Diffstat (limited to 'drivers/gpu/drm/i915/i915_gpu_error.c')
| -rw-r--r-- | drivers/gpu/drm/i915/i915_gpu_error.c | 116 |
1 files changed, 63 insertions, 53 deletions
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index ae84aa4b1467..9cd22cda17af 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -121,6 +121,7 @@ static void __i915_error_advance(struct drm_i915_error_state_buf *e, e->pos += len; } +__printf(2, 0) static void i915_error_vprintf(struct drm_i915_error_state_buf *e, const char *f, va_list args) { @@ -176,9 +177,14 @@ static void i915_error_puts(struct drm_i915_error_state_buf *e, #ifdef CONFIG_DRM_I915_COMPRESS_ERROR -static bool compress_init(struct z_stream_s *zstream) +struct compress { + struct z_stream_s zstream; + void *tmp; +}; + +static bool compress_init(struct compress *c) { - memset(zstream, 0, sizeof(*zstream)); + struct z_stream_s *zstream = memset(&c->zstream, 0, sizeof(c->zstream)); zstream->workspace = kmalloc(zlib_deflate_workspacesize(MAX_WBITS, MAX_MEM_LEVEL), @@ -191,14 +197,22 @@ static bool compress_init(struct z_stream_s *zstream) return false; } + c->tmp = NULL; + if (i915_has_memcpy_from_wc()) + c->tmp = (void *)__get_free_page(GFP_ATOMIC | __GFP_NOWARN); + return true; } -static int compress_page(struct z_stream_s *zstream, +static int compress_page(struct compress *c, void *src, struct drm_i915_error_object *dst) { + struct z_stream_s *zstream = &c->zstream; + zstream->next_in = src; + if (c->tmp && i915_memcpy_from_wc(c->tmp, src, PAGE_SIZE)) + zstream->next_in = c->tmp; zstream->avail_in = PAGE_SIZE; do { @@ -226,9 +240,11 @@ static int compress_page(struct z_stream_s *zstream, return 0; } -static void compress_fini(struct z_stream_s *zstream, +static void compress_fini(struct compress *c, struct drm_i915_error_object *dst) { + struct z_stream_s *zstream = &c->zstream; + if (dst) { zlib_deflate(zstream, Z_FINISH); dst->unused = zstream->avail_out; @@ -236,6 +252,9 @@ static void compress_fini(struct z_stream_s *zstream, zlib_deflateEnd(zstream); kfree(zstream->workspace); + + if (c->tmp) + free_page((unsigned long)c->tmp); } static void err_compression_marker(struct drm_i915_error_state_buf *m) @@ -245,28 +264,34 @@ static void err_compression_marker(struct drm_i915_error_state_buf *m) #else -static bool compress_init(struct z_stream_s *zstream) +struct compress { +}; + +static bool compress_init(struct compress *c) { return true; } -static int compress_page(struct z_stream_s *zstream, +static int compress_page(struct compress *c, void *src, struct drm_i915_error_object *dst) { unsigned long page; + void *ptr; page = __get_free_page(GFP_ATOMIC | __GFP_NOWARN); if (!page) return -ENOMEM; - dst->pages[dst->page_count++] = - memcpy((void *)page, src, PAGE_SIZE); + ptr = (void *)page; + if (!i915_memcpy_from_wc(ptr, src, PAGE_SIZE)) + memcpy(ptr, src, PAGE_SIZE); + dst->pages[dst->page_count++] = ptr; return 0; } -static void compress_fini(struct z_stream_s *zstream, +static void compress_fini(struct compress *c, struct drm_i915_error_object *dst) { } @@ -316,24 +341,6 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m, } } -static const char *hangcheck_action_to_str(enum intel_engine_hangcheck_action a) -{ - switch (a) { - case HANGCHECK_IDLE: - return "idle"; - case HANGCHECK_WAIT: - return "wait"; - case HANGCHECK_ACTIVE: - return "active"; - case HANGCHECK_KICK: - return "kick"; - case HANGCHECK_HUNG: - return "hung"; - } - - return "unknown"; -} - static void error_print_instdone(struct drm_i915_error_state_buf *m, struct drm_i915_error_engine *ee) { @@ -370,8 +377,8 @@ static void error_print_request(struct drm_i915_error_state_buf *m, if (!erq->seqno) return; - err_printf(m, "%s pid %d, seqno %8x:%08x, emitted %dms ago, head %08x, tail %08x\n", - prefix, erq->pid, + err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, emitted %dms ago, head %08x, tail %08x\n", + prefix, erq->pid, erq->ban_score, erq->context, erq->seqno, jiffies_to_msecs(jiffies - erq->jiffies), erq->head, erq->tail); @@ -441,9 +448,13 @@ static void error_print_engine(struct drm_i915_error_state_buf *m, err_printf(m, " waiting: %s\n", yesno(ee->waiting)); err_printf(m, " ring->head: 0x%08x\n", ee->cpu_ring_head); err_printf(m, " ring->tail: 0x%08x\n", ee->cpu_ring_tail); - err_printf(m, " hangcheck: %s [%d]\n", - hangcheck_action_to_str(ee->hangcheck_action), - ee->hangcheck_score); + err_printf(m, " hangcheck stall: %s\n", yesno(ee->hangcheck_stalled)); + err_printf(m, " hangcheck action: %s\n", + hangcheck_action_to_str(ee->hangcheck_action)); + err_printf(m, " hangcheck action timestamp: %lu, %u ms ago\n", + ee->hangcheck_timestamp, + jiffies_to_msecs(jiffies - ee->hangcheck_timestamp)); + error_print_request(m, " ELSP[0]: ", &ee->execlist[0]); error_print_request(m, " ELSP[1]: ", &ee->execlist[1]); } @@ -528,11 +539,10 @@ static void err_print_capabilities(struct drm_i915_error_state_buf *m, int i915_error_state_to_str(struct drm_i915_error_state_buf *m, const struct i915_error_state_file_priv *error_priv) { - struct drm_i915_private *dev_priv = to_i915(error_priv->dev); + struct drm_i915_private *dev_priv = error_priv->i915; struct pci_dev *pdev = dev_priv->drm.pdev; struct drm_i915_error_state *error = error_priv->error; struct drm_i915_error_object *obj; - int max_hangcheck_score; int i, j; if (!error) { @@ -549,22 +559,20 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, err_printf(m, "Uptime: %ld s %ld us\n", error->uptime.tv_sec, error->uptime.tv_usec); err_print_capabilities(m, &error->device_info); - max_hangcheck_score = 0; - for (i = 0; i < ARRAY_SIZE(error->engine); i++) { - if (error->engine[i].hangcheck_score > max_hangcheck_score) - max_hangcheck_score = error->engine[i].hangcheck_score; - } + for (i = 0; i < ARRAY_SIZE(error->engine); i++) { - if (error->engine[i].hangcheck_score == max_hangcheck_score && + if (error->engine[i].hangcheck_stalled && error->engine[i].pid != -1) { - err_printf(m, "Active process (on ring %s): %s [%d]\n", + err_printf(m, "Active process (on ring %s): %s [%d], context bans %d\n", engine_str(i), error->engine[i].comm, - error->engine[i].pid); + error->engine[i].pid, + error->engine[i].context_bans); } } err_printf(m, "Reset count: %u\n", error->reset_count); err_printf(m, "Suspend count: %u\n", error->suspend_count); + err_printf(m, "Platform: %s\n", intel_platform_name(error->device_info.platform)); err_printf(m, "PCI ID: 0x%04x\n", pdev->device); err_printf(m, "PCI Revision: 0x%02x\n", pdev->revision); err_printf(m, "PCI Subsystem: %04x:%04x\n", @@ -651,9 +659,10 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, if (obj) { err_puts(m, dev_priv->engine[i]->name); if (ee->pid != -1) - err_printf(m, " (submitted by %s [%d])", + err_printf(m, " (submitted by %s [%d], bans %d)", ee->comm, - ee->pid); + ee->pid, + ee->context_bans); err_printf(m, " --- gtt_offset = 0x%08x %08x\n", upper_32_bits(obj->gtt_offset), lower_32_bits(obj->gtt_offset)); @@ -801,7 +810,7 @@ i915_error_object_create(struct drm_i915_private *i915, struct i915_ggtt *ggtt = &i915->ggtt; const u64 slot = ggtt->error_capture.start; struct drm_i915_error_object *dst; - struct z_stream_s zstream; + struct compress compress; unsigned long num_pages; struct sgt_iter iter; dma_addr_t dma; @@ -821,7 +830,7 @@ i915_error_object_create(struct drm_i915_private *i915, dst->page_count = 0; dst->unused = 0; - if (!compress_init(&zstream)) { + if (!compress_init(&compress)) { kfree(dst); return NULL; } @@ -834,7 +843,7 @@ i915_error_object_create(struct drm_i915_private *i915, I915_CACHE_NONE, 0); s = io_mapping_map_atomic_wc(&ggtt->mappable, slot); - ret = compress_page(&zstream, (void __force *)s, dst); + ret = compress_page(&compress, (void __force *)s, dst); io_mapping_unmap_atomic(s); if (ret) @@ -849,7 +858,7 @@ unwind: dst = NULL; out: - compress_fini(&zstream, dst); + compress_fini(&compress, dst); ggtt->base.clear_range(&ggtt->base, slot, PAGE_SIZE); return dst; } @@ -941,7 +950,7 @@ static uint32_t i915_error_generate_code(struct drm_i915_private *dev_priv, * strictly a client bug. Use instdone to differentiate those some. */ for (i = 0; i < I915_NUM_ENGINES; i++) { - if (error->engine[i].hangcheck_action == HANGCHECK_HUNG) { + if (error->engine[i].hangcheck_stalled) { if (engine_id) *engine_id = i; @@ -1159,8 +1168,9 @@ static void error_record_engine_registers(struct drm_i915_error_state *error, ee->hws = I915_READ(mmio); } - ee->hangcheck_score = engine->hangcheck.score; + ee->hangcheck_timestamp = engine->hangcheck.action_timestamp; ee->hangcheck_action = engine->hangcheck.action; + ee->hangcheck_stalled = engine->hangcheck.stalled; if (USES_PPGTT(dev_priv)) { int i; @@ -1188,6 +1198,7 @@ static void record_request(struct drm_i915_gem_request *request, struct drm_i915_error_request *erq) { erq->context = request->ctx->hw_id; + erq->ban_score = request->ctx->ban_score; erq->seqno = request->global_seqno; erq->jiffies = request->emitted_jiffies; erq->head = request->head; @@ -1321,7 +1332,7 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv, } error->simulated |= - request->ctx->flags & CONTEXT_NO_ERROR_CAPTURE; + i915_gem_context_no_error_capture(request->ctx); ee->rq_head = request->head; ee->rq_post = request->postfix; @@ -1659,9 +1670,8 @@ void i915_error_state_put(struct i915_error_state_file_priv *error_priv) kref_put(&error_priv->error->ref, i915_error_state_free); } -void i915_destroy_error_state(struct drm_device *dev) +void i915_destroy_error_state(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = to_i915(dev); struct drm_i915_error_state *error; spin_lock_irq(&dev_priv->gpu_error.lock); |