MLBEDSW-8261: Fix regression on AvgPool

- When adding extended stride support for CONV_2D a
regression was introduced for AvgPool causing an
output diff for a particular test case.
- The reason was that the logic for forcing the
zero point to zero when generating the cmd stream
did not have a check for explicit padding.
- Updated logic to also include check for explicit
padding.

Change-Id: Iee4893a83a05279e592fe230f4d66d9c9ddb3e05
Signed-off-by: Johan Alfven <johan.alfven@arm.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/ethosu/vela/high_level_command_to_npu_op.py b/ethosu/vela/high_level_command_to_npu_op.py
index 4384f2c..5e9dffa 100644
--- a/ethosu/vela/high_level_command_to_npu_op.py
+++ b/ethosu/vela/high_level_command_to_npu_op.py
@@ -311,7 +311,7 @@ def use_zero_point_0(ps, tens: Tensor, is_ifm_tensor: bool) -> bool:
         if (
             ps.primary_op.original_type == Op.AvgPool
             and ps.primary_op.type == Op.Conv2DBias
-            and ps.primary_op.attrs.get("padding", None) == Padding.VALID
+            and ps.primary_op.attrs.get("padding", None) in (Padding.EXPLICIT, Padding.VALID)
         ):
             # Force zero point to 0 for AveragePool operators converted to a Conv2DBias with rounding away from
             # zero.