The image output using gst-launch-1.0 -e nvv4l2camerasrc has stripes

hi nvidia :

use this command :gst-launch-1.0 -e nvv4l2camerasrc device=/dev/video1 ! ‘video/x-raw(memory:NVMM),width=2880,height=1920,format=UYVY,framerate=30/1’
! nvvidconv ! ‘video/x-raw(memory:NVMM),format=NV12’ ! nvv4l2h264enc bitrate=20000000 ! h264parse ! qtmux ! filesink location=output0.mp4

The output0.mp4 has been successfully generated,but the video stream is striped

use this command is ok:gst-launch-1.0 -e v4l2src device=/dev/video1
! ‘video/x-raw,width=2880,height=1920,format=UYVY,framerate=30/1’
! nvvidconv
! ‘video/x-raw(memory:NVMM),format=NV12’
! nvv4l2h264enc bitrate=20000000
! h264parse ! qtmux ! filesink location=output1.mp4

What is the reason for this?Thank you very much. I would be grateful if you could respond

this is ok

Try set the preferred_stride by below command.

v4l2-ctl -c preferred_stride=2304

The problem is the same

Is it a resolution issue? I previously used 2560x1536 and it was fine

I think the root cause is need the output size(width) be 256 alignment.

Thanks

Only agx orin or thor need the output size(width) be 256 alignment?

My bad you should try below.

v4l2-ctl -c preferred_stride=3072

The problem is the same

Could you confirm by xvimagesink or nveglglessink?

What’s the version?

cat /etc/nv_tegra_release

R36 (release), REVISION: 4.4, GCID: 41062509, BOARD: generic, EABI: aarch64, DATE: Mon Jun 16 16:07:13 UTC 2025

KERNEL_VARIANT: oot

TARGET_USERSPACE_LIB_DIR=nvidia
TARGET_USERSPACE_LIB_DIR_PATH=usr/lib/aarch64-linux-gnu/nvidia

Hi,

Please download the source code from Jetson Linux Release 36.4.4 | NVIDIA Developer and apply the patch to check whether the issue exists.

r36_4_4_TEST_nvv4l2camera.zip (4.0 KB)

Thanks

cannot download r36_4_4_TEST_nvv4l2camera.zip

Hi,

Provide the patch as below.

diff --git a/gst-nvv4l2camera/Makefile b/gst-nvv4l2camera/Makefile
index 3dab045..4815208 100644
--- a/gst-nvv4l2camera/Makefile
+++ b/gst-nvv4l2camera/Makefile
@@ -1,4 +1,5 @@
-# Copyright (c) 2019-2022, NVIDIA CORPORATION. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2019-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: BSD-3-Clause
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions
@@ -30,15 +31,37 @@
 
 TARGET_DEVICE = $(shell gcc -dumpmachine | cut -f1 -d -)
 
+ifdef CUDA_VER
+CUDA_PATH := /usr/local/cuda-$(CUDA_VER)
+else
+CUDA_PATH := /usr/local/cuda
+endif
+
+# Location of the target rootfs
+ifeq ($(shell uname -m), aarch64)
+TARGET_ROOTFS :=
+else
+ifeq ($(TARGET_ROOTFS),)
+$(error Please specify the target rootfs path if you are cross-compiling)
+endif
+endif
+
+ALGO_CUDA_DIR=cuda
+
 GST_INSTALL_DIR?=/usr/lib/aarch64-linux-gnu/gstreamer-1.0/
 LIB_INSTALL_DIR?=/usr/lib/aarch64-linux-gnu/tegra/
 CFLAGS:=
-LIBS:= -lnvbufsurface -lnvbufsurftransform
+LIBS:= -lnvbufsurface -lnvbufsurftransform \
+	-lcuda -lcudart \
+	-L"$(TARGET_ROOTFS)/$(CUDA_PATH)/lib64" \
 
 
 SRCS := $(wildcard *.cpp)
 
-INCLUDES += -I./ -I../
+INCLUDES += -I./ -I../ -I./cuda \
+	-I/usr/src/jetson_multimedia_api/include/ \
+	-I$(CUDA_PATH)/include
+
 
 PKGS := gstreamer-1.0 \
 	gstreamer-base-1.0 \
@@ -48,6 +71,8 @@
 	libv4l2
 
 OBJS := $(SRCS:.cpp=.o)
+OBJS += $(ALGO_CUDA_DIR)/NvCudaProc.o
+
 
 CFLAGS += -fPIC \
 	-DEXPLICITLY_ADDED=1 \
@@ -66,6 +91,9 @@
 %.o: %.cpp
 	$(CC) -c $< $(CFLAGS) $(INCLUDES) -o $@
 
+$(ALGO_CUDA_DIR)/%.o: $(ALGO_CUDA_DIR)/%.cpp
+	$(CC) -c $< $(CFLAGS) $(INCLUDES) -o $@
+
 $(SO_NAME): $(OBJS)
 	$(CC) -shared -o $(SO_NAME) $(OBJS) $(LIBS) $(LDFLAGS)
 
diff --git a/gst-nvv4l2camera/cuda/NvCudaProc.cpp b/gst-nvv4l2camera/cuda/NvCudaProc.cpp
new file mode 100644
index 0000000..1c4d198
--- /dev/null
+++ b/gst-nvv4l2camera/cuda/NvCudaProc.cpp
@@ -0,0 +1,173 @@
+/*
+ * SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ * SPDX-License-Identifier: BSD-3-Clause
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *  * Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  * Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *  * Neither the name of NVIDIA CORPORATION nor the names of its
+ *    contributors may be used to endorse or promote products derived
+ *    from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
+ * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+ * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
+ * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <stdio.h>
+#include <cuda_runtime_api.h>
+#include "cudaEGL.h"
+#include "NvCudaProc.h"
+
+static void
+Handle_EGLImageMapping(EGLImageKHR srcimage,
+                       EGLImageKHR dstimage,
+                       int width,
+                       int height,
+                       int bytesPerPix);
+
+static void
+doMapping(CUdeviceptr srcDevPtr,
+          CUdeviceptr dstDevPtr,
+          int pitch,
+          int width,
+          int height,
+          int bytesPerPix);
+
+void
+HandleEGLImageMapping(void *pSrcEGLImage,
+                      void *pDstEGLImage,
+                      int width,
+                      int height,
+                      int bytesPerPix)
+{
+    EGLImageKHR *pSrcImage = (EGLImageKHR *)pSrcEGLImage;
+    EGLImageKHR *pDstImage = (EGLImageKHR *)pDstEGLImage;
+    Handle_EGLImageMapping(pSrcImage, pDstImage, width, height, bytesPerPix);
+}
+
+static void
+Handle_EGLImageMapping(EGLImageKHR srcimage,
+                       EGLImageKHR dstimage,
+                       int width,
+                       int height,
+                       int bytesPerPix)
+{
+    CUresult status;
+
+    CUeglFrame srcEglFrame;
+    CUgraphicsResource pSrcResource = NULL;
+    CUeglFrame dstEglFrame;
+    CUgraphicsResource pDstResource = NULL;
+
+    cudaFree(0);
+    status = cuGraphicsEGLRegisterImage(&pSrcResource, srcimage,
+                CU_GRAPHICS_MAP_RESOURCE_FLAGS_NONE);
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuGraphicsEGLRegisterImage failed: %d, cuda process stop\n",
+                        status);
+        return;
+    }
+
+    status = cuGraphicsEGLRegisterImage(&pDstResource, dstimage,
+                CU_GRAPHICS_MAP_RESOURCE_FLAGS_NONE);
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuGraphicsEGLRegisterImage failed: %d, cuda process stop\n",
+                        status);
+        return;
+    }
+
+    status = cuGraphicsResourceGetMappedEglFrame(&srcEglFrame, pSrcResource, 0, 0);
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuGraphicsSubResourceGetMappedArray failed\n");
+    }
+
+    status = cuGraphicsResourceGetMappedEglFrame(&dstEglFrame, pDstResource, 0, 0);
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuGraphicsSubResourceGetMappedArray failed\n");
+    }
+
+    status = cuCtxSynchronize();
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuCtxSynchronize failed\n");
+    }
+
+    if (srcEglFrame.frameType == CU_EGL_FRAME_TYPE_PITCH)
+    {
+        doMapping((CUdeviceptr) srcEglFrame.frame.pPitch[0],
+                  (CUdeviceptr) dstEglFrame.frame.pPitch[0],
+                  dstEglFrame.pitch,
+                  width,
+                  height,
+                  bytesPerPix);
+    }
+
+    status = cuCtxSynchronize();
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuCtxSynchronize failed after memcpy\n");
+    }
+
+    status = cuGraphicsUnregisterResource(pSrcResource);
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuGraphicsEGLUnRegisterResource failed: %d\n", status);
+    }
+
+    status = cuGraphicsUnregisterResource(pDstResource);
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuGraphicsEGLUnRegisterResource failed: %d\n", status);
+    }
+}
+
+
+static void
+doMapping(CUdeviceptr srcDevPtr,
+          CUdeviceptr dstDevPtr,
+          int pitch,
+          int width,
+          int height,
+          int bytesPerPix)
+{
+    CUDA_MEMCPY2D copyParams = {0};
+    copyParams.srcMemoryType = CU_MEMORYTYPE_DEVICE;
+    copyParams.srcDevice     = srcDevPtr;
+    copyParams.srcPitch      = width * bytesPerPix;
+    copyParams.srcXInBytes   = 0;
+    copyParams.srcY          = 0;
+
+    copyParams.dstMemoryType = CU_MEMORYTYPE_DEVICE;
+    copyParams.dstDevice     = dstDevPtr;
+    copyParams.dstPitch      = pitch;
+    copyParams.dstXInBytes   = 0;
+    copyParams.dstY          = 0;
+
+    copyParams.WidthInBytes  = width * bytesPerPix;
+    copyParams.Height        = height;
+
+    // Perform the copy
+    CUresult status = cuMemcpy2D(&copyParams);
+    if (status != CUDA_SUCCESS)
+    {
+        printf("cuMemcpy2D failed: %d\n", status);
+    }
+}
diff --git a/gst-nvv4l2camera/cuda/NvCudaProc.h b/gst-nvv4l2camera/cuda/NvCudaProc.h
new file mode 100644
index 0000000..e65fcb9
--- /dev/null
+++ b/gst-nvv4l2camera/cuda/NvCudaProc.h
@@ -0,0 +1,48 @@
+/*
+* SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+* SPDX-License-Identifier: BSD-3-Clause
+*
+* Redistribution and use in source and binary forms, with or without
+* modification, are permitted provided that the following conditions
+* are met:
+*  * Redistributions of source code must retain the above copyright
+*    notice, this list of conditions and the following disclaimer.
+*  * Redistributions in binary form must reproduce the above copyright
+*    notice, this list of conditions and the following disclaimer in the
+*    documentation and/or other materials provided with the distribution.
+*  * Neither the name of NVIDIA CORPORATION nor the names of its
+*    contributors may be used to endorse or promote products derived
+*    from this software without specific prior written permission.
+*
+* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
+* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+* PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR
+* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
+* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+#ifndef __NVCUDAPROC_H
+#define __NVCUDAPROC_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* This function is called when padding insertion is needed for proper alignment.
+ * The function registers surfaces to EGL images for CUDA access and uses cuMemcpy2D
+ * to perform correct alignment.
+ * */
+void HandleEGLImageMapping(void *pSrcEGLImage, void *pDstEGLImage, int width, int height,
+                        int bytesPerPix);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif
diff --git a/gst-nvv4l2camera/gstnvv4l2camerasrc.cpp b/gst-nvv4l2camera/gstnvv4l2camerasrc.cpp
index b211f66..1a04f64 100644
--- a/gst-nvv4l2camera/gstnvv4l2camerasrc.cpp
+++ b/gst-nvv4l2camera/gstnvv4l2camerasrc.cpp
@@ -1,5 +1,7 @@
 /*
- * Copyright (c) 2019-2024, NVIDIA CORPORATION. All rights reserved.
+
+ * SPDX-FileCopyrightText: Copyright (c) 2019-2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ * SPDX-License-Identifier: BSD-3-Clause
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
@@ -55,6 +57,7 @@
 #include <fcntl.h>
 #include <linux/videodev2.h>
 #include <sys/ioctl.h>
+#include "NvCudaProc.h"
 
 /* Capture caps of plugin */
 #define CAPTURE_CAPS \
@@ -198,6 +201,19 @@
     goto error;
   }
 
+  if (nvbuf->surface->surfaceList[0].pitch !=
+        nvbuf->surface->surfaceList[0].planeParams.bytesPerPix[0] *
+        nvbuf->surface->surfaceList[0].width && !self->remap_surf) {
+    ret = NvBufSurfaceAllocate (&self->remap_surf, 1,  &param);
+
+    if (ret != 0) {
+      GST_ERROR ("%s: NvBufSurfaceAllocate Failed \n", __func__);
+      goto error;
+    }
+
+    self->remap_surf->numFilled = 1;
+  }
+
   nvbuf->surface->numFilled = 1;
 
   gst_memory_init (GST_MEMORY_CAST (mem), flags, allocator, NULL,
@@ -242,6 +258,23 @@
 
   GST_DEBUG ("%s:\n", __func__);
 
+  GstNVV4l2MemoryAllocator *nvmm_allocator = GST_NVMEMORY_ALLOCATOR (allocator);
+  if (!nvmm_allocator) {
+    GST_DEBUG ("%s: NULL GstNVV4l2MemoryAllocator \n", __func__);
+    return;
+  }
+
+  GstNvV4l2CameraSrc *self = (GstNvV4l2CameraSrc*) nvmm_allocator->owner;
+  if (self == NULL) {
+    GST_DEBUG ("%s: NULL NvV4l2CameraSrc \n", __func__);
+    return;
+  }
+
+  if (self->remap_surf) {
+    NvBufSurfaceDestroy(self->remap_surf);
+    self->remap_surf = NULL;
+  }
+
   GstNVV4l2Memory *nv_mem = (GstNVV4l2Memory *) mem;
   if (nv_mem == NULL)
     return;
@@ -376,6 +409,28 @@
     memcpy (nv_mem->nvcam_buf->buffer, &tmp_v4l2_buf, sizeof (struct v4l2_buffer));
   }
 
+  if(src->remap_surf) {
+    if (-1 == NvBufSurfaceMapEglImage(nv_mem->nvcam_buf->surface, 0)){
+      GST_ERROR_OBJECT(pool, "NvBufSurfaceMapEglImage failed");
+    }
+
+    if (-1 == NvBufSurfaceMapEglImage(src->remap_surf, 0)){
+      GST_ERROR_OBJECT(pool, "NvBufSurfaceMapEglImage failed");
+    }
+
+    HandleEGLImageMapping(nv_mem->nvcam_buf->surface->surfaceList[0].mappedAddr.eglImage,
+      src->remap_surf->surfaceList[0].mappedAddr.eglImage,
+      nv_mem->nvcam_buf->surface->surfaceList[0].width,
+      nv_mem->nvcam_buf->surface->surfaceList[0].height,
+      nv_mem->nvcam_buf->surface->surfaceList[0].planeParams.bytesPerPix[0]);
+
+    /* Destroy EGLImage */
+    NvBufSurfaceUnMapEglImage(nv_mem->nvcam_buf->surface, 0);
+    NvBufSurfaceUnMapEglImage(src->remap_surf, 0);
+
+    NvBufSurfaceCopy (src->remap_surf, nv_mem->nvcam_buf->surface);
+  }
+
   return ret;
 
 no_memblk:
diff --git a/gst-nvv4l2camera/gstnvv4l2camerasrc.hpp b/gst-nvv4l2camera/gstnvv4l2camerasrc.hpp
index e126dee..3eb80c7 100644
--- a/gst-nvv4l2camera/gstnvv4l2camerasrc.hpp
+++ b/gst-nvv4l2camera/gstnvv4l2camerasrc.hpp
@@ -1,5 +1,6 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION. All rights reserved.
+ * SPDX-FileCopyrightText: Copyright (c) 2019-2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ * SPDX-License-Identifier: BSD-3-Clause
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
@@ -78,6 +79,7 @@
   struct v4l2_capability *caps;
   struct v4l2_format *fmt;
   struct v4l2_requestbuffers *req;
+  NvBufSurface *remap_surf;
   gint index;
   guint cap_buf;

Thanks