Re: [Intel-gfx] [PATCH 2/5] i915/drm/guc: Don't pass CTB while writing

[Michal Wajdeczko]

On Thu, 16 Jan 2020 19:53:08 +0100, Daniele Ceraolo Spurio  
<daniele.ceraolospu...@intel.com> wrote:



On 1/15/20 6:08 AM, Michal Wajdeczko wrote:
Since we only have one SEND buffer we don't need to explicitly pass
it to the write function.
 Signed-off-by: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
  drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c | 13 +++++++------
  1 file changed, 7 insertions(+), 6 deletions(-)
 diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c  
b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
index 0d3556a820a3..dedbf3b8ab01 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
@@ -311,12 +311,13 @@ static u32 ct_get_next_fence(struct intel_guc_ct  
*ct)
   *                   ^-----------------len-------------------^
   */
  -static int ctb_write(struct intel_guc_ct_buffer *ctb,
-                    const u32 *action,
-                    u32 len /* in dwords */,
-                    u32 fence,
-                    bool want_response)
+static int ct_write(struct intel_guc_ct *ct,
+                   const u32 *action,
+                   u32 len /* in dwords */,
+                   u32 fence,
+                   bool want_response)
  {
+       struct intel_guc_ct_buffer *ctb = &ct->ctbs[CTB_SEND];
        struct guc_ct_buffer_desc *desc = ctb->desc;
        u32 head = desc->head / 4;   /* in dwords */
        u32 tail = desc->tail / 4;   /* in dwords */
@@ -492,7 +493,7 @@ static int ct_send(struct intel_guc_ct *ct,
        list_add_tail(&request.link, &ct->requests.pending);
        spin_unlock_irqrestore(&ct->requests.lock, flags);
  -     err = ctb_write(ctb, action, len, fence, !!response_buf);
+       err = ct_write(ct, action, len, fence, !!response_buf);

I'd update wait_for_ctb_desc_update() to work on struct intel_guc_ct as  
well, so we can hide the ctb desc access in the lower level functions  
and drop the ctb variable from ct_send(). With or without that:

in next series I was planning to remove wait_for_ctb_desc_update()
that's why I didn't update it now


Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>

Daniele

        if (unlikely(err))
                goto unlink;
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx