[Mesa-dev] [PATCH 6/6] i965: Add some batchbuffer debugging code.

[Kenneth Graunke]

On Tuesday, November 28, 2017 4:13:21 PM PST Kenneth Graunke wrote:
> Uncommenting these blocks stress tests the batch/state growth code by
> trying to "grow" a buffer to the same size at least once per batch.
> 
> When anything goes wrong with this code, dumping the validation list
> is a useful way to figure out what's happening.
> ---
>  src/mesa/drivers/dri/i965/intel_batchbuffer.c | 27 +++++++++++++++++++++++++++
>  1 file changed, 27 insertions(+)
> 
> diff --git a/src/mesa/drivers/dri/i965/intel_batchbuffer.c b/src/mesa/drivers/dri/i965/intel_batchbuffer.c
> index 12d165d7236..b5b9468ed65 100644
> --- a/src/mesa/drivers/dri/i965/intel_batchbuffer.c
> +++ b/src/mesa/drivers/dri/i965/intel_batchbuffer.c
> @@ -64,6 +64,21 @@
>  static void
>  intel_batchbuffer_reset(struct brw_context *brw);
>  
> +UNUSED static void
> +dump_validation_list(struct intel_batchbuffer *batch)
> +{
> +   fprintf(stderr, "Validation list (length %d):\n", batch->exec_count);
> +
> +   for (int i = 0; i < batch->exec_count; i++) {
> +      assert(batch->validation_list[i].handle ==
> +             batch->exec_bos[i]->gem_handle);
> +      fprintf(stderr, "[%d] = %d %s %p\n", i,
> +              batch->validation_list[i].handle,
> +              batch->exec_bos[i]->name,
> +              batch->exec_bos[i]);
> +   }
> +}
> +
>  static bool
>  uint_key_compare(const void *a, const void *b)
>  {
> @@ -996,6 +1011,18 @@ _intel_batchbuffer_flush_fence(struct brw_context *brw,
>     brw_finish_batch(brw);
>     intel_upload_finish(brw);
>  
> +   /* Test batch/state growing by swapping a BO for one of the same size */
> +   if (0) {
> +      const unsigned used = 4 * USED_BATCH(brw->batch);
> +      grow_buffer(brw, &brw->batch.batch, used, brw->batch.batch.bo->size);
> +      brw->batch.map_next = (void *) brw->batch.batch.map + used;
> +   }
> +   if (0) {

I've updated this locally to:

+   if (0 && brw->batch.state.bo->index != -1) {

because the buffer growing code assumes the BO is actually in the
validation list.  It normally is, unless you don't emit any state
at all, say for really simple batches.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: This is a digitally signed message part.
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20171129/84ffc563/attachment.sig>