The Linux Kernel Mailing List
 help / color / mirror / Atom feed
* [PATCH 1/2] accel: ethosu: Fix element size accounting for cmd stream validation
@ 2026-07-20 23:14 Rob Herring (Arm)
  2026-07-20 23:14 ` [PATCH 2/2] accel: ethosu: Handle U85 internal chaining buffer Rob Herring (Arm)
  2026-07-21 16:50 ` [PATCH 1/2] accel: ethosu: Fix element size accounting for cmd stream validation Tomeu Vizoso
  0 siblings, 2 replies; 4+ messages in thread
From: Rob Herring (Arm) @ 2026-07-20 23:14 UTC (permalink / raw)
  To: Tomeu Vizoso, Oded Gabbay, Thomas Zimmermann, Frank Li
  Cc: dri-devel, linux-kernel

There are 2 issues with the element size handling in the command stream
validation which result in too small of a size calculated when the
element size is 16/32/64 bits.

For NHWC format, the element size is simply missing from the
calculation.

The bitfield for the element size is different between IFM/IFM2 and
OFM. IFM and IFM2 encode the precision in parameter bits 2:3, while OFM
uses bits 1:2.

Fixes: 5a5e9c0228e6 ("accel: Add Arm Ethos-U NPU driver")
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
---
 drivers/accel/ethosu/ethosu_gem.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/accel/ethosu/ethosu_gem.c b/drivers/accel/ethosu/ethosu_gem.c
index 3401883e207f..1daff32610c7 100644
--- a/drivers/accel/ethosu/ethosu_gem.c
+++ b/drivers/accel/ethosu/ethosu_gem.c
@@ -194,7 +194,7 @@ static u64 dma_length(struct ethosu_validated_cmdstream_info *info,
 
 static u64 feat_matrix_length(struct ethosu_validated_cmdstream_info *info,
 			      struct feat_matrix *fm,
-			      u32 x, u32 y, u32 c)
+			      u32 x, u32 y, u32 c, bool ofm)
 {
 	u32 element_size, storage = fm->precision >> 14;
 	int tile = 0;
@@ -231,10 +231,11 @@ static u64 feat_matrix_length(struct ethosu_validated_cmdstream_info *info,
 
 	switch ((fm->precision >> 6) & 0x3) { // format
 	case 0: //nhwc:
-		addr += x * fm->stride_x + c;
+		element_size = BIT((fm->precision >> (ofm ? 1 : 2)) & 0x3);
+		addr += x * fm->stride_x + c * element_size;
 		break;
 	case 1: //nhcwb16:
-		element_size = BIT((fm->precision >> 1) & 0x3);
+		element_size = BIT((fm->precision >> (ofm ? 1 : 2)) & 0x3);
 
 		addr += (c / 16) * fm->stride_c + (16 * x + (c & 0xf)) * element_size;
 		break;
@@ -268,7 +269,7 @@ static int calc_sizes(struct drm_device *ddev,
 			return -EINVAL;
 
 		len = feat_matrix_length(info, &st->ifm, ifm_width,
-					 ifm_height, st->ifm.depth);
+					 ifm_height, st->ifm.depth, false);
 		dev_dbg(ddev->dev, "op %d: IFM:%d:0x%llx-0x%llx\n",
 			op, st->ifm.region, st->ifm.base[0], len);
 		if (len == U64_MAX)
@@ -277,7 +278,7 @@ static int calc_sizes(struct drm_device *ddev,
 
 	if (ifm2) {
 		len = feat_matrix_length(info, &st->ifm2, st->ifm.depth,
-					 0, st->ofm.depth);
+					 0, st->ofm.depth, false);
 		dev_dbg(ddev->dev, "op %d: IFM2:%d:0x%llx-0x%llx\n",
 			op, st->ifm2.region, st->ifm2.base[0], len);
 		if (len == U64_MAX)
@@ -309,7 +310,7 @@ static int calc_sizes(struct drm_device *ddev,
 	}
 
 	len = feat_matrix_length(info, &st->ofm, st->ofm.width,
-				 st->ofm.height[2], st->ofm.depth);
+				 st->ofm.height[2], st->ofm.depth, true);
 	dev_dbg(ddev->dev, "op %d: OFM:%d:0x%llx-0x%llx\n",
 		op, st->ofm.region, st->ofm.base[0], len);
 	if (len == U64_MAX)
@@ -333,7 +334,7 @@ static int calc_sizes_elemwise(struct drm_device *ddev,
 		depth = st->ifm.broadcast & 0x4 ? 0 : st->ofm.depth;
 
 		len = feat_matrix_length(info, &st->ifm, width,
-					 height, depth);
+					 height, depth, false);
 		dev_dbg(ddev->dev, "op %d: IFM:%d:0x%llx-0x%llx\n",
 			op, st->ifm.region, st->ifm.base[0], len);
 		if (len == U64_MAX)
@@ -346,7 +347,7 @@ static int calc_sizes_elemwise(struct drm_device *ddev,
 		depth = st->ifm2.broadcast & 0x4 ? 0 : st->ofm.depth;
 
 		len = feat_matrix_length(info, &st->ifm2, width,
-					 height, depth);
+					 height, depth, false);
 		dev_dbg(ddev->dev, "op %d: IFM2:%d:0x%llx-0x%llx\n",
 			op, st->ifm2.region, st->ifm2.base[0], len);
 		if (len == U64_MAX)
@@ -354,7 +355,7 @@ static int calc_sizes_elemwise(struct drm_device *ddev,
 	}
 
 	len = feat_matrix_length(info, &st->ofm, st->ofm.width,
-				 st->ofm.height[2], st->ofm.depth);
+				 st->ofm.height[2], st->ofm.depth, true);
 	dev_dbg(ddev->dev, "op %d: OFM:%d:0x%llx-0x%llx\n",
 		op, st->ofm.region, st->ofm.base[0], len);
 	if (len == U64_MAX)
-- 
2.53.0


^ permalink raw reply related	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2026-07-21 16:50 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-07-20 23:14 [PATCH 1/2] accel: ethosu: Fix element size accounting for cmd stream validation Rob Herring (Arm)
2026-07-20 23:14 ` [PATCH 2/2] accel: ethosu: Handle U85 internal chaining buffer Rob Herring (Arm)
2026-07-21 16:49   ` Tomeu Vizoso
2026-07-21 16:50 ` [PATCH 1/2] accel: ethosu: Fix element size accounting for cmd stream validation Tomeu Vizoso

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox