Skip to contentSkip to conversation

linux-wandboard: Update kernel to 3.10.53-1.1.0_ga

PATCH In review Alfonso Tames 3 participants 14 messages Feb 9, 2015 – Feb 11, 2015

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 linux-wandboard: Update kernel to 3.10.53-1.1.0_ga

From: Alfonso Tamés <[email protected]>

This updates the Wandboards' kernel to 3.10.53-1.1.0_ga

---
 conf/machine/include/wandboard.inc                 |   2 +-
 .../linux/linux-wandboard-3.10.31/defconfig        | 402 ---------------------
 .../linux/linux-wandboard-3.10.53/defconfig        | 402 +++++++++++++++++++++
 recipes-kernel/linux/linux-wandboard_3.10.31.bb    |  15 -
 recipes-kernel/linux/linux-wandboard_3.10.53.bb    |  15 +
 5 files changed, 418 insertions(+), 418 deletions(-)
 delete mode 100644 recipes-kernel/linux/linux-wandboard-3.10.31/defconfig
 create mode 100644 recipes-kernel/linux/linux-wandboard-3.10.53/defconfig
 delete mode 100644 recipes-kernel/linux/linux-wandboard_3.10.31.bb
 create mode 100644 recipes-kernel/linux/linux-wandboard_3.10.53.bb
conf/machine/include/wandboard.inc +1 -1
index 8629c3b..9dffb43 100644
@@ -4,7 +4,7 @@ include conf/machine/include/imx-base.inc
 include conf/machine/include/tune-cortexa9.inc
 
 PREFERRED_PROVIDER_virtual/kernel ?= "linux-wandboard"
-PREFERRED_VERSION_linux-wandboard ?= "3.10.31"
+PREFERRED_VERSION_linux-wandboard ?= "3.10.53"
 
 SERIAL_CONSOLE = "115200 ttymxc0"
 
recipes-kernel/linux/linux-wandboard-3.10.31/defconfig deleted +0 -402

402 lines not on this page · raw message

recipes-kernel/linux/linux-wandboard-3.10.53/defconfig new file +402 -0

402 lines not on this page · raw message

recipes-kernel/linux/linux-wandboard_3.10.31.bb deleted +0 -15
deleted file mode 100644
index 94e433e..0000000
@@ -1,15 +0,0 @@
-# adapted from linux-imx.inc, copyright (C) 2012-2014 O.S. Systems Software LTDA
-# Released under the MIT license (see COPYING.MIT for the terms)
-
-include linux-wandboard.inc
-require recipes-kernel/linux/linux-dtb.inc
-
-DEPENDS += "lzop-native bc-native"
-
-# Wandboard branch - based on 3.10.31_1.1.0_beta2 from Freescale git
-
-SRCBRANCH = "wandboard_imx_3.10.31_1.1.0_beta2"
-SRCREV = "a63c7c073c385bdc9d041c9795a18754f4544dba"
-LOCALVERSION = "_1.1.0_beta2-wandboard"
-
-COMPATIBLE_MACHINE = "(wandboard)"
recipes-kernel/linux/linux-wandboard_3.10.53.bb new file +15 -0
new file mode 100644
index 0000000..ce76274
@@ -0,0 +1,15 @@
+# adapted from linux-imx.inc, copyright (C) 2012-2014 O.S. Systems Software LTDA
+# Released under the MIT license (see COPYING.MIT for the terms)
+
+include linux-wandboard.inc
+require recipes-kernel/linux/linux-dtb.inc
+
+DEPENDS += "lzop-native bc-native"
+
+# Wandboard branch - based on 3.10.53_1.1.0_ga from Freescale git
+
+SRCBRANCH = "wandboard_imx_3.10.53_1.1.0_ga"
+SRCREV = "13bb18444208d5065f0e6345c7743e9662bc97d2"
+LOCALVERSION = "_1.1.0_ga-wandboard"
+
+COMPATIBLE_MACHINE = "(wandboard)"
Signature · 1 line
2.1.0
#1 lore raw
Quoted 18 lines · Alfonso Tames

From: Alfonso Tamés <[email protected]>

This updates the Wandboards' kernel to 3.10.53-1.1.0_ga

---
 conf/machine/include/wandboard.inc                 |   2 +-
 .../linux/linux-wandboard-3.10.31/defconfig        | 402
---------------------
 .../linux/linux-wandboard-3.10.53/defconfig        | 402
+++++++++++++++++++++
 recipes-kernel/linux/linux-wandboard_3.10.31.bb    |  15 -
 recipes-kernel/linux/linux-wandboard_3.10.53.bb    |  15 +
 5 files changed, 418 insertions(+), 418 deletions(-)
 delete mode 100644 recipes-kernel/linux/linux-wandboard-3.10.31/defconfig
 create mode 100644 recipes-kernel/linux/linux-wandboard-3.10.53/defconfig
 delete mode 100644 recipes-kernel/linux/linux-wandboard_3.10.31.bb
 create mode 100644 recipes-kernel/linux/linux-wandboard_3.10.53.bb

Looks good to me, Alfonso. I've applied it and tested it with wandcam and gstreamer-1.0. Looks like I'm getting some sketchy problem that results in a segfault when a gstreamer pipeline shuts down, but not sure if that is due to this kernel or if it existed in previous kernels (3.10.17 versions or 3.10.31).

#2 lore raw

John, Can you please also test with Gst1.0 (the new gstreamer 1.0 from freescale)?

Thanks!

Quoted 22 lines · John Weber
Quoted 18 lines · Alfonso Tames

From: Alfonso Tamés <[email protected]>

This updates the Wandboards' kernel to 3.10.53-1.1.0_ga

---
 conf/machine/include/wandboard.inc                 |   2 +-
 .../linux/linux-wandboard-3.10.31/defconfig        | 402 ---------------------
 .../linux/linux-wandboard-3.10.53/defconfig        | 402 +++++++++++++++++++++
 recipes-kernel/linux/linux-wandboard_3.10.31.bb    |  15 -
 recipes-kernel/linux/linux-wandboard_3.10.53.bb    |  15 +
 5 files changed, 418 insertions(+), 418 deletions(-)
 delete mode 100644 recipes-kernel/linux/linux-wandboard-3.10.31/defconfig
 create mode 100644 recipes-kernel/linux/linux-wandboard-3.10.53/defconfig
 delete mode 100644 recipes-kernel/linux/linux-wandboard_3.10.31.bb
 create mode 100644 recipes-kernel/linux/linux-wandboard_3.10.53.bb

Looks good to me, Alfonso. I've applied it and tested it with wandcam and gstreamer-1.0. Looks like I'm getting some sketchy problem that results in a segfault when a gstreamer pipeline shuts down, but not sure if that is due to this kernel or if it existed in previous kernels (3.10.17 versions or 3.10.31).

#3 lore raw

On 2/10/15, 3:08 PM, Alfonso Tamés wrote:
John, Can you please also test with Gst1.0 (the new gstreamer 1.0 from freescale)?

Thanks!
I can.  I'd rather use (and help with) Carlos's gstreamer work, but I'll give the fsl stuff a try.  I'll try a fresh install as well.

#4 lore raw
Quoted 4 lines

I can. I'd rather use (and help with) Carlos's gstreamer work, but I'll give the fsl stuff a try. I'll try a fresh install as well.

I understand, but there is a strange issue with Carlos’s gstreamer 1.0 with Wandboard (see https://github.com/Freescale/gstreamer-imx/issues/15 <https://github.com/Freescale/gstreamer-imx/issues/15>). On the other hand, Freescale’s gstreamer 1.0 is producing video tearing with the Wandcam. It would be good to know if any other board is experiencing this as well.

Regards,

Alfonso

#5 lore raw
In reply to John Weber

Thanks John, when I first tested it in an existing Yocto installation, Kodi 14.1 was producing choppy video. I generated a new image from a fresh Yocto install and has been working flawlessly in Kodi for several days.

Can you make a fresh install?

Thanks,

Alfonso

Quoted 22 lines · John Weber
Quoted 18 lines · Alfonso Tames

From: Alfonso Tamés <[email protected]>

This updates the Wandboards' kernel to 3.10.53-1.1.0_ga

---
 conf/machine/include/wandboard.inc                 |   2 +-
 .../linux/linux-wandboard-3.10.31/defconfig        | 402 ---------------------
 .../linux/linux-wandboard-3.10.53/defconfig        | 402 +++++++++++++++++++++
 recipes-kernel/linux/linux-wandboard_3.10.31.bb    |  15 -
 recipes-kernel/linux/linux-wandboard_3.10.53.bb    |  15 +
 5 files changed, 418 insertions(+), 418 deletions(-)
 delete mode 100644 recipes-kernel/linux/linux-wandboard-3.10.31/defconfig
 create mode 100644 recipes-kernel/linux/linux-wandboard-3.10.53/defconfig
 delete mode 100644 recipes-kernel/linux/linux-wandboard_3.10.31.bb
 create mode 100644 recipes-kernel/linux/linux-wandboard_3.10.53.bb

Looks good to me, Alfonso. I've applied it and tested it with wandcam and gstreamer-1.0. Looks like I'm getting some sketchy problem that results in a segfault when a gstreamer pipeline shuts down, but not sure if that is due to this kernel or if it existed in previous kernels (3.10.17 versions or 3.10.31).

#6 lore raw
In reply to John Weber
Quoted 4 lines

Looks good to me, Alfonso. I've applied it and tested it with wandcam and gstreamer-1.0. Looks like I'm getting some sketchy problem that results in a segfault when a gstreamer pipeline shuts down, but not sure if that is due to this kernel or if it existed in previous kernels (3.10.17 versions or 3.10.31).

:(

I just tested a fresh install and I'm getting the same error:

root@wandboard-quad:~# GST_DEBUG=*:3; gst-launch-1.0 -e imxv4l2src capture-mode=4 fps-n=30 ! imxvpuenc_h264 bitrate=4096 ! h264parse ! mp4mux ! filesink location=video.mp4
...
EOS received - stopping pipeline...
Execution ended after 0:00:04.088140334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c72498 in waitpid () from /lib/libpthread.so.0
#1  0x76c9cb64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()
Spinning.  Please run 'gdb gst-launch-1.0 1011' to continue debugging, Ctrl-C to quit, or Ctrl-\ to dump core.
^C

Any ideas Otavio?

#7 lore raw
Quoted 29 lines · Alfonso Tamés
Quoted 9 lines

Looks good to me, Alfonso. I've applied it and tested it with wandcam and gstreamer-1.0. Looks like I'm getting some sketchy problem that results in a segfault when a gstreamer pipeline shuts down, but not sure if that is due to this kernel or if it existed in previous kernels (3.10.17 versions or 3.10.31).

:(

I just tested a fresh install and I'm getting the same error:

root@wandboard-quad:~# GST_DEBUG=*:3; gst-launch-1.0 -e imxv4l2src
capture-mode=4 fps-n=30 ! imxvpuenc_h264 bitrate=4096 ! h264parse ! mp4mux
! filesink location=video.mp4
...
EOS received - stopping pipeline...
Execution ended after 0:00:04.088140334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c72498 in waitpid () from /lib/libpthread.so.0
#1  0x76c9cb64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()
Spinning.  Please run 'gdb gst-launch-1.0 1011' to continue debugging,
Ctrl-C to quit, or Ctrl-\ to dump core.
^C
Quoted 2 lines

Any ideas Otavio?

Can you see this with a videotestsrc?

#8 lore raw

On 2/11/15, 5:02 AM, John Weber wrote:

On Tuesday, February 10, 2015, Alfonso Tamés <[email protected]> wrote:

Looks good to me, Alfonso.  I've applied it and tested it with wandcam and gstreamer-1.0.  Looks like I'm getting some sketchy problem that results in a segfault when a gstreamer pipeline shuts down, but not sure if that is due to this kernel or if it existed in previous kernels (3.10.17 versions or 3.10.31).

:(

I just tested a fresh install and I'm getting the same error:

root@wandboard-quad:~# GST_DEBUG=*:3; gst-launch-1.0 -e imxv4l2src capture-mode=4 fps-n=30 ! imxvpuenc_h264 bitrate=4096 ! h264parse ! mp4mux ! filesink location=video.mp4
 ...
EOS received - stopping pipeline...
Execution ended after 0:00:04.088140334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c72498 in waitpid () from /lib/libpthread.so.0
#1  0x76c9cb64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()
Spinning.  Please run 'gdb gst-launch-1.0 1011' to continue debugging, Ctrl-C to quit, or Ctrl-\ to dump core.
 ^C

Any ideas Otavio?

Can you see this with a videotestsrc?
I'm seeing the same issue (segfault) on this side with a really fresh build of fsl-image-multimedia-full using master, using Carlos's gstreamer-1.0 plugins:

gst-launch-1.0 imxv4l2src capture-mode=4 fps-n=30 ! imxipuvideosink

I'm not seeing a problem with this pipeline, which is from an image with the gst1.0-fsl-plugins installed (note that Freescale's element has the same name as Carlos's, but uses caps to set the format instead of properties):
gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxv4l2sink

I do see tearing with the same pipeline, but using the imxipuvideosink instead of the imxv4l2sink.

The pipeline (similar to yours but using the gst1.0-fsl-plugin imxv4l2src) works for me:
gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxvpuenc_h264 bitrate=2000 ! h264parse ! mp4mux ! filesink location=video.mp4

#9 lore raw
Quoted 6 lines

The pipeline (similar to yours but using the gst1.0-fsl-plugin imxv4l2src) works for me:
gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxvpuenc_h264 bitrate=2000 ! h264parse ! mp4mux ! filesink location=video.mp4

Are you sure, that pipeline segfaults immediately for me:

root@wandboard-quad:~# gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxvpuenc_h264 bitrate=2000 ! h264parse ! mp4mux ! filesink location=video.mp4
.....
New clock: GstSystemClock
ERROR: from element /GstPipeline:pipeline0/GstImxV4l2Src:imxv4l2src0: Internal data flow error.
Additional debug info:
/opt/yocto/master_egl/build/tmp/work/cortexa9hf-vfp-neon-poky-linux-gnueabi/gstreamer1.0/1.4.5-r0/gstreamer-1.4.5/libs/gst/base/gstbasesrc.c(2933): gst_base_src_loop (): /GstPipeline:pipeline0/GstImxV4l2Src:imxv4l2src0:
streaming task paused, reason not-negotiated (-4)
Execution ended after 0:00:00.093394334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c89498 in waitpid () from /lib/libpthread.so.0
#1  0x76cb3b64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()
#10 lore raw
Quoted 6 lines · Alfonso Tamés
Quoted 4 lines

The pipeline (similar to yours but using the gst1.0-fsl-plugin imxv4l2src) works for me:
gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxvpuenc_h264 bitrate=2000 ! h264parse ! mp4mux ! filesink location=video.mp4

Are you sure, that pipeline segfaults immediately for me:

Quoted 16 lines
root@wandboard-quad:~# gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxvpuenc_h264 bitrate=2000 ! h264parse ! mp4mux ! filesink location=video.mp4
.....
New clock: GstSystemClock
ERROR: from element /GstPipeline:pipeline0/GstImxV4l2Src:imxv4l2src0: Internal data flow error.
Additional debug info:
/opt/yocto/master_egl/build/tmp/work/cortexa9hf-vfp-neon-poky-linux-gnueabi/gstreamer1.0/1.4.5-r0/gstreamer-1.4.5/libs/gst/base/gstbasesrc.c(2933): gst_base_src_loop (): /GstPipeline:pipeline0/GstImxV4l2Src:imxv4l2src0:
streaming task paused, reason not-negotiated (-4)
Execution ended after 0:00:00.093394334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c89498 in waitpid () from /lib/libpthread.so.0
#1  0x76cb3b64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()

It does on my end when I use imxv4l2src from Carlos's gstreamer plugins. It does not when use the plugin of the same name from gst1.0-fsl-plugin.

John

#11 lore raw

Oh, of course, you mentioned it! My bad.

If you are using a Wandboard Quad, could you test this pipeline using Carlos's gstreamer1.0?

gst-launch-1.0 videotestsrc ! imxvpuenc_h264 ! h264parse ! imxvpudec ! imxeglvivsink

This was causing timeouts in the quads few seconds after start but seemed to work fine in the duals.

Thanks!

Quoted 28 lines · John Weber
Quoted 4 lines · Alfonso Tamés
Quoted 2 lines

The pipeline (similar to yours but using the gst1.0-fsl-plugin imxv4l2src) works for me:
gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxvpuenc_h264 bitrate=2000 ! h264parse ! mp4mux ! filesink location=video.mp4

Are you sure, that pipeline segfaults immediately for me:

Quoted 14 lines
root@wandboard-quad:~# gst-launch-1.0 imxv4l2src ! video/x-raw,format=I420,width=1280,height=720,framerate=30/1 ! imxvpuenc_h264 bitrate=2000 ! h264parse ! mp4mux ! filesink location=video.mp4
.....
New clock: GstSystemClock
ERROR: from element /GstPipeline:pipeline0/GstImxV4l2Src:imxv4l2src0: Internal data flow error.
Additional debug info:
/opt/yocto/master_egl/build/tmp/work/cortexa9hf-vfp-neon-poky-linux-gnueabi/gstreamer1.0/1.4.5-r0/gstreamer-1.4.5/libs/gst/base/gstbasesrc.c(2933): gst_base_src_loop (): /GstPipeline:pipeline0/GstImxV4l2Src:imxv4l2src0:
streaming task paused, reason not-negotiated (-4)
Execution ended after 0:00:00.093394334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c89498 in waitpid () from /lib/libpthread.so.0
#1  0x76cb3b64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()

It does on my end when I use imxv4l2src from Carlos's gstreamer plugins. It does not when use the plugin of the same name from gst1.0-fsl-plugin.

John
It does on my end when I use imxv4l2src from Carlos's gstreamer plugins. It does not when use the plugin of the same name from gst1.0-fsl-plugin.

John

#12 lore raw
In reply to John Weber
Quoted 14 lines
root@wandboard-quad:~# GST_DEBUG=*:3; gst-launch-1.0 -e imxv4l2src capture-mode=4 fps-n=30 ! imxvpuenc_h264 bitrate=4096 ! h264parse ! mp4mux ! filesink location=video.mp4
...
EOS received - stopping pipeline...
Execution ended after 0:00:04.088140334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c72498 in waitpid () from /lib/libpthread.so.0
#1  0x76c9cb64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()
Spinning.  Please run 'gdb gst-launch-1.0 1011' to continue debugging, Ctrl-C to quit, or Ctrl-\ to dump core.
^C
Quoted 3 lines

Any ideas Otavio?

Can you see this with a videotestsrc?

With videotestsrc it works fine:

GST_DEBUG=*:3; gst-launch-1.0 -e videotestsrc ! imxvpuenc_h264 bitrate=4096 ! h264parse ! mp4mux ! filesink location=video.mp4
...
^Chandling interrupt.
Interrupt: Stopping pipeline ...
EOS on shutdown enabled -- Forcing EOS on the pipeline
Waiting for EOS...
Got EOS from element "pipeline0".
EOS received - stopping pipeline...
Execution ended after 0:00:04.654726001
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Setting pipeline to NULL ...
Freeing pipeline ...
root@wandboard-quad:~#

It seems the problem is located at imxv4l2src

#13 lore raw
Quoted 42 lines · Alfonso Tamés
Quoted 17 lines
root@wandboard-quad:~# GST_DEBUG=*:3; gst-launch-1.0 -e imxv4l2src
capture-mode=4 fps-n=30 ! imxvpuenc_h264 bitrate=4096 ! h264parse ! mp4mux !
filesink location=video.mp4
...
EOS received - stopping pipeline...
Execution ended after 0:00:04.088140334
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Caught SIGSEGV
#0  0x76c72498 in waitpid () from /lib/libpthread.so.0
#1  0x76c9cb64 in g_on_error_stack_trace () from /usr/lib/libglib-2.0.so.0
#2  0x0000b4e0 in ?? ()
Spinning.  Please run 'gdb gst-launch-1.0 1011' to continue debugging,
Ctrl-C to quit, or Ctrl-\ to dump core.
^C
Quoted 1 line

Any ideas Otavio?

Can you see this with a videotestsrc?

With videotestsrc it works fine:

GST_DEBUG=*:3; gst-launch-1.0 -e videotestsrc ! imxvpuenc_h264 bitrate=4096 ! h264parse ! mp4mux ! filesink location=video.mp4
...
^Chandling interrupt.
Interrupt: Stopping pipeline ...
EOS on shutdown enabled -- Forcing EOS on the pipeline
Waiting for EOS...
Got EOS from element "pipeline0".
EOS received - stopping pipeline...
Execution ended after 0:00:04.654726001
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Setting pipeline to NULL ...
Freeing pipeline ...
root@wandboard-quad:~#

Quoted 1 line

It seems the problem is located at imxv4l2src

I know Carlos is preparing an update; check master branch of the plugin to check if it fixes your issue.

Signature · 3 lines
Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750