| Message ID | 20210906225636.14683-30-laurent.pinchart@ideasonboard.com |
|---|---|
| State | Accepted |
| Headers | show |
| Series |
|
| Related | show |
On 06/09/2021 23:56, Laurent Pinchart wrote: > The V4L2 compatibility layer supports the single-planar API only, and > thuss expose a single V4L2 buffer plane to applications, regardless of s/thuss/thus/ s/expose/exposes/ > the number of planes in the FrameBuffer. For multi-planar frame buffers, > the bytesused value isn't correct as it only takes the first plane into > account. Fix it by summing the bytesused values for all FrameBuffer > planes. > > Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > --- > src/v4l2/v4l2_camera_proxy.cpp | 7 ++++++- > 1 file changed, 6 insertions(+), 1 deletion(-) > > diff --git a/src/v4l2/v4l2_camera_proxy.cpp b/src/v4l2/v4l2_camera_proxy.cpp > index 07d2250bb846..aeffef26c5bf 100644 > --- a/src/v4l2/v4l2_camera_proxy.cpp > +++ b/src/v4l2/v4l2_camera_proxy.cpp > @@ -11,6 +11,7 @@ > #include <array> > #include <errno.h> > #include <linux/videodev2.h> > +#include <numeric> > #include <set> > #include <string.h> > #include <sys/mman.h> > @@ -211,7 +212,11 @@ void V4L2CameraProxy::updateBuffers() > > switch (fmd.status) { > case FrameMetadata::FrameSuccess: > - buf.bytesused = fmd.planes()[0].bytesused; > + buf.bytesused = std::accumulate(fmd.planes().begin(), > + fmd.planes().end(), 0, > + [](unsigned int total, const auto &plane) { lambda's always look like black magic to me. I assume total is either automatically initialised, or initialised by the 0 after fmd.planes().end() Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> > + return total + plane.bytesused; > + }); > buf.field = V4L2_FIELD_NONE; > buf.timestamp.tv_sec = fmd.timestamp / 1000000000; > buf.timestamp.tv_usec = fmd.timestamp % 1000000; >
Hi Laurent, On 07/09/2021 16:17, Kieran Bingham wrote: > On 06/09/2021 23:56, Laurent Pinchart wrote: >> The V4L2 compatibility layer supports the single-planar API only, and >> thuss expose a single V4L2 buffer plane to applications, regardless of > > s/thuss/thus/ > s/expose/exposes/ > >> the number of planes in the FrameBuffer. For multi-planar frame buffers, >> the bytesused value isn't correct as it only takes the first plane into >> account. Fix it by summing the bytesused values for all FrameBuffer >> planes. >> >> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> >> --- >> src/v4l2/v4l2_camera_proxy.cpp | 7 ++++++- >> 1 file changed, 6 insertions(+), 1 deletion(-) >> >> diff --git a/src/v4l2/v4l2_camera_proxy.cpp b/src/v4l2/v4l2_camera_proxy.cpp >> index 07d2250bb846..aeffef26c5bf 100644 >> --- a/src/v4l2/v4l2_camera_proxy.cpp >> +++ b/src/v4l2/v4l2_camera_proxy.cpp >> @@ -11,6 +11,7 @@ >> #include <array> >> #include <errno.h> >> #include <linux/videodev2.h> >> +#include <numeric> >> #include <set> >> #include <string.h> >> #include <sys/mman.h> >> @@ -211,7 +212,11 @@ void V4L2CameraProxy::updateBuffers() >> >> switch (fmd.status) { >> case FrameMetadata::FrameSuccess: >> - buf.bytesused = fmd.planes()[0].bytesused; >> + buf.bytesused = std::accumulate(fmd.planes().begin(), >> + fmd.planes().end(), 0, >> + [](unsigned int total, const auto &plane) { > > lambda's always look like black magic to me. > I assume total is either automatically initialised, or initialised by > the 0 after fmd.planes().end() I read back cpp reference for that one :-), and yes, total takes the value of init, which is 0. > > Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Reviewed-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com> > >> + return total + plane.bytesused; >> + }); >> buf.field = V4L2_FIELD_NONE; >> buf.timestamp.tv_sec = fmd.timestamp / 1000000000; >> buf.timestamp.tv_usec = fmd.timestamp % 1000000; >>
Hi Kieran, On Tue, Sep 07, 2021 at 03:17:48PM +0100, Kieran Bingham wrote: > On 06/09/2021 23:56, Laurent Pinchart wrote: > > The V4L2 compatibility layer supports the single-planar API only, and > > thuss expose a single V4L2 buffer plane to applications, regardless of > > s/thuss/thus/ > s/expose/exposes/ > > > the number of planes in the FrameBuffer. For multi-planar frame buffers, > > the bytesused value isn't correct as it only takes the first plane into > > account. Fix it by summing the bytesused values for all FrameBuffer > > planes. > > > > Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > > --- > > src/v4l2/v4l2_camera_proxy.cpp | 7 ++++++- > > 1 file changed, 6 insertions(+), 1 deletion(-) > > > > diff --git a/src/v4l2/v4l2_camera_proxy.cpp b/src/v4l2/v4l2_camera_proxy.cpp > > index 07d2250bb846..aeffef26c5bf 100644 > > --- a/src/v4l2/v4l2_camera_proxy.cpp > > +++ b/src/v4l2/v4l2_camera_proxy.cpp > > @@ -11,6 +11,7 @@ > > #include <array> > > #include <errno.h> > > #include <linux/videodev2.h> > > +#include <numeric> > > #include <set> > > #include <string.h> > > #include <sys/mman.h> > > @@ -211,7 +212,11 @@ void V4L2CameraProxy::updateBuffers() > > > > switch (fmd.status) { > > case FrameMetadata::FrameSuccess: > > - buf.bytesused = fmd.planes()[0].bytesused; > > + buf.bytesused = std::accumulate(fmd.planes().begin(), > > + fmd.planes().end(), 0, > > + [](unsigned int total, const auto &plane) { > > lambda's always look like black magic to me. > I assume total is either automatically initialised, or initialised by > the 0 after fmd.planes().end() The latter (https://en.cppreference.com/w/cpp/algorithm/accumulate). > Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com> > > > + return total + plane.bytesused; > > + }); > > buf.field = V4L2_FIELD_NONE; > > buf.timestamp.tv_sec = fmd.timestamp / 1000000000; > > buf.timestamp.tv_usec = fmd.timestamp % 1000000;
diff --git a/src/v4l2/v4l2_camera_proxy.cpp b/src/v4l2/v4l2_camera_proxy.cpp index 07d2250bb846..aeffef26c5bf 100644 --- a/src/v4l2/v4l2_camera_proxy.cpp +++ b/src/v4l2/v4l2_camera_proxy.cpp @@ -11,6 +11,7 @@ #include <array> #include <errno.h> #include <linux/videodev2.h> +#include <numeric> #include <set> #include <string.h> #include <sys/mman.h> @@ -211,7 +212,11 @@ void V4L2CameraProxy::updateBuffers() switch (fmd.status) { case FrameMetadata::FrameSuccess: - buf.bytesused = fmd.planes()[0].bytesused; + buf.bytesused = std::accumulate(fmd.planes().begin(), + fmd.planes().end(), 0, + [](unsigned int total, const auto &plane) { + return total + plane.bytesused; + }); buf.field = V4L2_FIELD_NONE; buf.timestamp.tv_sec = fmd.timestamp / 1000000000; buf.timestamp.tv_usec = fmd.timestamp % 1000000;
The V4L2 compatibility layer supports the single-planar API only, and thuss expose a single V4L2 buffer plane to applications, regardless of the number of planes in the FrameBuffer. For multi-planar frame buffers, the bytesused value isn't correct as it only takes the first plane into account. Fix it by summing the bytesused values for all FrameBuffer planes. Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> --- src/v4l2/v4l2_camera_proxy.cpp | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)