[Mesa-dev] [PATCH 3/6] i965: Enable hardware-generated binding tables on render path.

[Abdiel Janulgue]

On 06/26/2015 10:46 AM, Chris Wilson wrote:
> On Fri, Jun 26, 2015 at 08:51:59AM +0300, Abdiel Janulgue wrote:
>> This patch implements the binding table enable command which is also
>> used to allocate a binding table pool where where hardware-generated
>> binding table entries are flushed into. Each binding table offset in
>> the binding table pool is unique per each shader stage that are
>> enabled within a batch.
>>
>> Also insert the required brw_tracked_state objects to enable
>> hw-generated binding tables in normal render path.
>>
>> v2: - Use MOCS in binding table pool alloc for GEN8
>>     - Fix spurious offset when allocating binding table pool entry
>>       and start from zero instead.
>> v3  - Include GEN8 fix for spurious offset above.
>>
>> Cc: kenneth at whitecape.org
>> Signed-off-by: Abdiel Janulgue <abdiel.janulgue at linux.intel.com>
>> ---
>>  src/mesa/drivers/dri/i965/brw_binding_tables.c | 87 ++++++++++++++++++++++++++
>>  src/mesa/drivers/dri/i965/brw_context.c        |  4 ++
>>  src/mesa/drivers/dri/i965/brw_context.h        |  6 ++
>>  src/mesa/drivers/dri/i965/brw_state.h          |  6 ++
>>  src/mesa/drivers/dri/i965/brw_state_upload.c   |  4 ++
>>  src/mesa/drivers/dri/i965/gen7_disable.c       |  4 +-
>>  src/mesa/drivers/dri/i965/gen8_disable.c       |  4 +-
>>  src/mesa/drivers/dri/i965/intel_batchbuffer.c  |  4 ++
>>  8 files changed, 115 insertions(+), 4 deletions(-)
>>
>> diff --git a/src/mesa/drivers/dri/i965/brw_binding_tables.c b/src/mesa/drivers/dri/i965/brw_binding_tables.c
>> index 98ff0dd..6bc540f 100644
>> --- a/src/mesa/drivers/dri/i965/brw_binding_tables.c
>> +++ b/src/mesa/drivers/dri/i965/brw_binding_tables.c
>> @@ -170,6 +170,93 @@ const struct brw_tracked_state brw_gs_binding_table = {
>>     .emit = brw_gs_upload_binding_table,
>>  };
>>  
>> +/**
>> + * Hardware-generated binding tables for the resource streamer
>> + */
>> +void
>> +gen7_disable_hw_binding_tables(struct brw_context *brw)
>> +{
>> +   int pkt_len = brw->gen >= 8 ? 4 : 3;
> 
> How do you anticipate this function being used? As it looks like
> a if (!brw->resource_streamer_active) return; would make the function
> foolproof.
> -Chris

This seems to be a valid call even with RS not around as long as
3DSTATE_BINDING_TABLE_POOL_ALLOC is set to disable. HW didn't complain.
But I guess you're right, it's safer to skip it out without RS. I'll do
that in a follow-up patch.

-Abdiel