[meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Neena Busireddy]

From: Neena Busireddy <neena.busireddy@freescale.com>

This version is released with 3.14.52-1.1.0_ga release.

Includes many of the bug fixes and stability improvements.
For more information refer to i.MX Linux Release Notes from following link:
http://www.nxp.com/products/microcontrollers-and-processors/arm-processors/i.mx-applications-processors-based-on-arm-cores/i.mx-6-processors/i.mx6qp/i.mx-6quad-processors-high-performance-3d-graphics-hd-video-arm-cortex-a9-core:i.MX6Q?fpsp=1&tab=Documentation_Tab#nogo

L3.14.52_1.1.0_LINUX_DOCS package is under Supporting Information.

Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
---
 ...{imx-gpu-viv_5.0.11.p7.1-hfp.bb => imx-gpu-viv_5.0.11.p7.4-hfp.bb} | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
 rename recipes-graphics/imx-gpu-viv/{imx-gpu-viv_5.0.11.p7.1-hfp.bb => imx-gpu-viv_5.0.11.p7.4-hfp.bb} (54%)

diff --git a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
similarity index 54%
rename from recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
rename to recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
index 22f097b..7895b6a 100644
--- a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
+++ b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
@@ -3,7 +3,7 @@
 
 require imx-gpu-viv.inc
 
-SRC_URI[md5sum] = "b26211f322c325e76f912f432989f2ee"
-SRC_URI[sha256sum] = "e49ca1d31ca83bb50f6672325db37ae3754eb9073ad98a37790ab1209ea5cba8"
+SRC_URI[md5sum] = "d348def4c050e215efe970828752e5b9"
+SRC_URI[sha256sum] = "b80838104f2f2e3b0baf2f1aacdca263c1c92589ce919b3be40aaad23f686b7f"
 
 PACKAGE_FP_TYPE = "hardfp"
-- 
1.9.1

[meta-fsl-arm][PATCH v3 2/5] imx-gpu-viv: Upgrade sfp to 5.0.11.p7.4 version

[Neena Busireddy]

From: Neena Busireddy <neena.busireddy@freescale.com>

Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
---
 ...{imx-gpu-viv_5.0.11.p7.1-sfp.bb => imx-gpu-viv_5.0.11.p7.4-sfp.bb} | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
 rename recipes-graphics/imx-gpu-viv/{imx-gpu-viv_5.0.11.p7.1-sfp.bb => imx-gpu-viv_5.0.11.p7.4-sfp.bb} (63%)

diff --git a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-sfp.bb b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-sfp.bb
similarity index 63%
rename from recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-sfp.bb
rename to recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-sfp.bb
index 610f7ce..ace3d5f 100644
--- a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-sfp.bb
+++ b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-sfp.bb
@@ -3,8 +3,8 @@
 
 require imx-gpu-viv.inc
 
-SRC_URI[md5sum] = "5dbe194e87e4092379b195e866cdb687"
-SRC_URI[sha256sum] = "43fea6ebbb1222e0f594fff3b48147f4bb6173b85f2caa9fa6bf04a1c85ee93a"
+SRC_URI[md5sum] = "6423b4a803106e7a3fff0911daaff410"
+SRC_URI[sha256sum] = "3ea201f91ce9b56455e4983ab1055e8175415fbbf2800dc097d47e78ceefba34"
 
 # FIXME skip the QA error for viv-samples
 INSANE_SKIP_${PN} += "rpaths"
-- 
1.9.1

[meta-fsl-arm][PATCH v3 3/5] kernel-module-imx-gpu-viv: Upgrade to 5.0.11.p7.4 version

[Neena Busireddy]

From: Neena Busireddy <neena.busireddy@freescale.com>

Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
---
 ....0.11.p7.1+fslc.bb => kernel-module-imx-gpu-viv_5.0.11.p7.4+fslc.bb} | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
 rename recipes-kernel/kernel-modules/{kernel-module-imx-gpu-viv_5.0.11.p7.1+fslc.bb => kernel-module-imx-gpu-viv_5.0.11.p7.4+fslc.bb} (91%)

diff --git a/recipes-kernel/kernel-modules/kernel-module-imx-gpu-viv_5.0.11.p7.1+fslc.bb b/recipes-kernel/kernel-modules/kernel-module-imx-gpu-viv_5.0.11.p7.4+fslc.bb
similarity index 91%
rename from recipes-kernel/kernel-modules/kernel-module-imx-gpu-viv_5.0.11.p7.1+fslc.bb
rename to recipes-kernel/kernel-modules/kernel-module-imx-gpu-viv_5.0.11.p7.4+fslc.bb
index 0504cfa..3d545ed 100644
--- a/recipes-kernel/kernel-modules/kernel-module-imx-gpu-viv_5.0.11.p7.1+fslc.bb
+++ b/recipes-kernel/kernel-modules/kernel-module-imx-gpu-viv_5.0.11.p7.4+fslc.bb
@@ -9,7 +9,7 @@ LIC_FILES_CHKSUM = "file://COPYING;md5=12f884d2ae1ff87c09e5b7ccc2c4ca7e"
 
 PV .= "+git${SRCPV}"
 
-SRCREV = "eeeb23c0fb1cee01318088d417025263479c44ac"
+SRCREV = "2b3290d1c1ce383e860a0fd6b53074848a848e65"
 SRC_URI = "git://github.com/Freescale/kernel-module-imx-gpu-viv.git;protocol=https"
 
 S = "${WORKDIR}/git"
-- 
1.9.1

[meta-fsl-arm][PATCH v3 4/5] weston: Add patches to fix bugs

[Neena Busireddy]

From: Neena Busireddy <neena.busireddy@freescale.com>

- Patch to fix setenv and clear environments.
- Patch to fix screen blur for Qt5 CinematicExperience.

Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
---
 ...GS-1235-Fix-setenv-and-clear-environments.patch | 41 ++++++++++
 ...x-for-Qt5_CinematicExperience-will-meet-s.patch | 93 ++++++++++++++++++++++
 recipes-graphics/wayland/weston_%.bbappend         |  5 +-
 3 files changed, 138 insertions(+), 1 deletion(-)
 create mode 100644 recipes-graphics/wayland/weston/0004-MGS-1235-Fix-setenv-and-clear-environments.patch
 create mode 100644 recipes-graphics/wayland/weston/0005-MGS-1252-Fix-for-Qt5_CinematicExperience-will-meet-s.patch

diff --git a/recipes-graphics/wayland/weston/0004-MGS-1235-Fix-setenv-and-clear-environments.patch b/recipes-graphics/wayland/weston/0004-MGS-1235-Fix-setenv-and-clear-environments.patch
new file mode 100644
index 0000000..e4f8cd9
--- /dev/null
+++ b/recipes-graphics/wayland/weston/0004-MGS-1235-Fix-setenv-and-clear-environments.patch
@@ -0,0 +1,41 @@
+From d09d0595b472d6bae35e78272fc670d494f6e408 Mon Sep 17 00:00:00 2001
+From: Prabhu <prabhu.sundararaj@freescale.com>
+Date: Tue, 17 Nov 2015 22:00:42 -0600
+Subject: [PATCH] MGS-1235 : Fix setenv and clear environments
+
+When setenv is set with parameter null causing crash.
+Fixed with unsetenv to restore to previous state
+
+Date: Nov 17, 2015
+Signed-off-by: Prabhu <prabhu.sundararaj@freescale.com>
+
+-1235
+---
+ src/compositor-fbdev.c | 11 ++++++++++-
+ 1 file changed, 10 insertions(+), 1 deletion(-)
+
+diff --git a/src/compositor-fbdev.c b/src/compositor-fbdev.c
+index ed25576..c6b8e5a 100644
+--- a/src/compositor-fbdev.c
++++ b/src/compositor-fbdev.c
+@@ -598,7 +598,16 @@ fbdev_output_create(struct fbdev_backend *backend,
+ 			fprintf(stderr, "failed to create window\n");
+ 			 return 0;
+ 		}
+-		setenv("FB_FRAMEBUFFER_0", fbenv, 1);
++		/* restore the previous value*/
++		if(fbenv != NULL)
++		{
++			setenv("FB_FRAMEBUFFER_0", fbenv, 1);
++		}
++		else
++		{
++			unsetenv("FB_FRAMEBUFFER_0");
++		}
++
+ 
+ 		if (gal2d_renderer->output_create(&output->base,
+ 					output->display,
+-- 
+2.5.1
+
diff --git a/recipes-graphics/wayland/weston/0005-MGS-1252-Fix-for-Qt5_CinematicExperience-will-meet-s.patch b/recipes-graphics/wayland/weston/0005-MGS-1252-Fix-for-Qt5_CinematicExperience-will-meet-s.patch
new file mode 100644
index 0000000..82d3798
--- /dev/null
+++ b/recipes-graphics/wayland/weston/0005-MGS-1252-Fix-for-Qt5_CinematicExperience-will-meet-s.patch
@@ -0,0 +1,93 @@
+From 515b1f1c1902ae22914a93982935e2328d2240d2 Mon Sep 17 00:00:00 2001
+From: Yong Gan <b45748@freescale.com>
+Date: Thu, 19 Nov 2015 15:30:24 +0800
+Subject: [PATCH 2/2] MGS-1252: Fix for Qt5_CinematicExperience will meet
+ screen blur
+
+When the Qt5_CinematicExperience is created, window creates with 1920x1080
+and the starting coordiantes more than screen coordinate.
+Ths cause access beyond the surface and system hang on lcd display
+
+Date: Nov 18, 2015
+Signed-off-by: Prabhu <prabhu.sundararaj@freescale.com>
+---
+ src/gal2d-renderer.c | 30 ++++++++++++++++++------------
+ 1 file changed, 18 insertions(+), 12 deletions(-)
+
+diff --git a/src/gal2d-renderer.c b/src/gal2d-renderer.c
+index c68f02c..6e7d640 100644
+--- a/src/gal2d-renderer.c
++++ b/src/gal2d-renderer.c
+@@ -661,12 +661,11 @@ repaint_region(struct weston_view *ev, struct weston_output *output, struct gal2
+ 	gcsRECT srcRect = {0};
+ 	gcsRECT dstrect = {0};
+ 	gctUINT32 horFactor, verFactor;
+-	int useStretch =1;
+ 	int useFilterBlit = 0;
+ 	gctUINT srcWidth = 0;
+ 	gctUINT srcHeight = 0;
+ 	gctUINT32 srcStride[3];
+-	gceSURF_FORMAT srcFormat;;
++	gceSURF_FORMAT srcFormat;
+ 	gctUINT32 srcPhyAddr[3];
+ 	gctUINT32 dstPhyAddr[3];
+ 	gctUINT dstWidth = 0;
+@@ -702,14 +701,16 @@ repaint_region(struct weston_view *ev, struct weston_output *output, struct gal2
+ 	srcRect.right = ev->surface->width;
+ 	srcRect.bottom = ev->surface->height;
+ 
+-	if(useFilterBlit)
++	dstsurface = go->nNumBuffers > 1 ?
++					go->renderSurf[go->activebuffer] :
++					go->offscreenSurface;
++	gcmVERIFY_OK(gcoSURF_GetAlignedSize(dstsurface, &dstWidth, &dstHeight, (gctINT *)&dstStrides));
++	gcmVERIFY_OK(gcoSURF_Lock(dstsurface, &dstPhyAddr[0], gcvNULL));
++	gcmVERIFY_OK(gcoSURF_Unlock(dstsurface, gcvNULL));
++
++	if(galIsYUVFormat(srcFormat) == gcvSTATUS_TRUE)
+ 	{
+-		dstsurface = go->nNumBuffers > 1 ?
+-						go->renderSurf[go->activebuffer] :
+-						go->offscreenSurface;
+-		gcmVERIFY_OK(gcoSURF_GetAlignedSize(dstsurface, &dstWidth, &dstHeight, (gctINT *)&dstStrides));
+-		gcmVERIFY_OK(gcoSURF_Lock(dstsurface, &dstPhyAddr[0], gcvNULL));
+-		gcmVERIFY_OK(gcoSURF_Unlock(dstsurface, gcvNULL));
++		useFilterBlit = 1;
+ 	}
+ 	else
+ 	{
+@@ -739,7 +740,7 @@ repaint_region(struct weston_view *ev, struct weston_output *output, struct gal2
+ 		dstrect.right = bb_rects[0].x2;
+ 		dstrect.bottom = bb_rects[0].y2;
+ 
+-		if(dstrect.right < 0 || dstrect.bottom < 0)
++		if(dstrect.right < 0 || dstrect.bottom < 0 || dstrect.left > dstWidth || dstrect.top > dstHeight)
+ 		{
+ 			break;
+ 		}
+@@ -829,8 +830,7 @@ repaint_region(struct weston_view *ev, struct weston_output *output, struct gal2
+ 			}
+ 			else
+ 			{
+-				if(useStretch)
+-					gcmVERIFY_OK(galGetStretchFactors(&srcRect, &dstrect, &horFactor, &verFactor));
++				gcmVERIFY_OK(galGetStretchFactors(&srcRect, &dstrect, &horFactor, &verFactor));
+ 
+ 				if(verFactor == 65536 && horFactor == 65536)
+ 				{
+@@ -839,6 +839,12 @@ repaint_region(struct weston_view *ev, struct weston_output *output, struct gal2
+ 				}
+ 				else
+ 				{
++					dstrect.right = dstrect.right < dstWidth ? dstrect.right : dstWidth;
++					dstrect.bottom = dstrect.bottom < dstHeight ? dstrect.bottom : dstHeight;
++					srcRect.right = srcRect.right < dstWidth ? srcRect.right : dstWidth;
++					srcRect.bottom = srcRect.bottom < dstHeight ? srcRect.bottom : dstHeight;
++
++					gcmVERIFY_OK(galGetStretchFactors(&srcRect, &dstrect, &horFactor, &verFactor));
+ 					/* Program the stretch factors. */
+ 					gcmVERIFY_OK(gco2D_SetStretchFactors(gr->gcoEngine2d, horFactor, verFactor));
+ 
+-- 
+1.9.1
+
diff --git a/recipes-graphics/wayland/weston_%.bbappend b/recipes-graphics/wayland/weston_%.bbappend
index a7496da..6c8ba6d 100644
--- a/recipes-graphics/wayland/weston_%.bbappend
+++ b/recipes-graphics/wayland/weston_%.bbappend
@@ -1,6 +1,9 @@
 FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
 
-SRC_URI_append_mx6 = " file://0001-MGS-840-Add-i.MX6-support-for-weston.patch"
+SRC_URI_append_mx6 = " \
+    file://0001-MGS-840-Add-i.MX6-support-for-weston.patch \
+    file://0004-MGS-1235-Fix-setenv-and-clear-environments.patch \
+    file://0005-MGS-1252-Fix-for-Qt5_CinematicExperience-will-meet-s.patch"
 
 PACKAGECONFIG_append_mx6q = " cairo-glesv2"
 PACKAGECONFIG_append_mx6dl = " cairo-glesv2"
-- 
1.9.1

[meta-fsl-arm][PATCH v3 5/5] xf86-video-imxfb-vivante: Upgrade to 5.0.11.p7.4 version

[Neena Busireddy]

From: Neena Busireddy <neena.busireddy@freescale.com>

- License changed to MIT.
- Remove the upstreamed patch.

Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
---
 .../xorg-driver/xf86-video-imxfb-vivante.inc       |   2 +-
 ...-EXA-Warnings-causes-GCC5.2-build-to-fail.patch | 137 ---------------------
 ....bb => xf86-video-imxfb-vivante_5.0.11.p7.4.bb} |   7 +-
 3 files changed, 4 insertions(+), 142 deletions(-)
 delete mode 100644 recipes-graphics/xorg-driver/xf86-video-imxfb-vivante/0001-MGS-907-X11-EXA-Warnings-causes-GCC5.2-build-to-fail.patch
 rename recipes-graphics/xorg-driver/{xf86-video-imxfb-vivante_5.0.11.p7.1.bb => xf86-video-imxfb-vivante_5.0.11.p7.4.bb} (48%)

diff --git a/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante.inc b/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante.inc
index 9f856b3..aa37555 100644
--- a/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante.inc
+++ b/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante.inc
@@ -10,7 +10,7 @@ inherit autotools-brokensep update-rc.d pkgconfig
 
 DEPENDS += "virtual/xserver virtual/libx11 virtual/libgal-x11 imx-gpu-viv pixman"
 
-LIC_FILES_CHKSUM = "file://COPYING-GPL-2;md5=59530bdf33659b29e73d4adb9f9f6552"
+LIC_FILES_CHKSUM = "file://COPYING-MIT;md5=b5e9d9f5c02ea831ab3ecf802bb7c4f3"
 
 SRC_URI = "${FSL_MIRROR}/xserver-xorg-video-imx-viv-${PV}.tar.gz \
             file://rc.autohdmi"
diff --git a/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante/0001-MGS-907-X11-EXA-Warnings-causes-GCC5.2-build-to-fail.patch b/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante/0001-MGS-907-X11-EXA-Warnings-causes-GCC5.2-build-to-fail.patch
deleted file mode 100644
index cb7e432..0000000
--- a/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante/0001-MGS-907-X11-EXA-Warnings-causes-GCC5.2-build-to-fail.patch
+++ /dev/null
@@ -1,137 +0,0 @@
-From 194e357faf3d7433ad4f1f2cf46f79d314ea5b38 Mon Sep 17 00:00:00 2001
-From: Prabhu <prabhu.sundararaj@freescale.com>
-Date: Wed, 29 Jul 2015 15:32:01 -0500
-Subject: [PATCH] MGS-907: X11 EXA: Warnings causes GCC5.2 build to fail
-
-VIV2DGPUSurfaceReAllocNonCached and imxRefreshModes were not
-declared causing compiler warning.
-
-Upstream Status: Pending
-
-Date: July 29, 2015
-Signed-off-by: Prabhu <prabhu.sundararaj@freescale.com>
----
- EXA/src/vivante_extension/vivante_ext.c      | 4 ++--
- EXA/src/vivante_fbdev/imx_display.c          | 4 +++-
- EXA/src/vivante_fbdev/vivante_fbdev_driver.c | 2 ++
- EXA/src/vivante_gal/vivante_gal.h            | 1 +
- EXA/src/vivante_gal/vivante_gal_surface.c    | 3 ++-
- EXA/src/vivante_gal/vivante_priv.h           | 4 ++++
- 6 files changed, 14 insertions(+), 4 deletions(-)
- mode change 100755 => 100644 EXA/src/vivante_extension/vivante_ext.c
- mode change 100755 => 100644 EXA/src/vivante_gal/vivante_gal_surface.c
-
-diff --git a/EXA/src/vivante_extension/vivante_ext.c b/EXA/src/vivante_extension/vivante_ext.c
-old mode 100755
-new mode 100644
-index 4bcc5ba..0d1085b
---- a/EXA/src/vivante_extension/vivante_ext.c
-+++ b/EXA/src/vivante_extension/vivante_ext.c
-@@ -54,6 +54,7 @@
- #include "vivante.h"
- #include "vivante_priv.h"
- #include "vivante_common.h"
-+#include "imx_display.h"
- 
- static unsigned char VIVEXTReqCode = 0;
- static int VIVEXTErrorBase;
-@@ -162,8 +163,7 @@ static int ProcVIVEXTDrawableSetFlag(register ClientPtr client)
-             VivPtr pViv = VIVPTR_FROM_SCREEN(pScreen);
-             if(pViv)
-             {
--                VIVGPUPtr gpuctx = (VIVGPUPtr) pViv->mGrCtx.mGpu;
--                VIV2DGPUSurfaceReAllocNonCached(gpuctx, ppriv);
-+                VIV2DGPUSurfaceReAllocNonCached(&pViv->mGrCtx, ppriv);
-                 // set tiled if asked
-                 if(stuff->flag & VIVPIXMAP_FLAG_SHARED_TILED)
-                 {
-diff --git a/EXA/src/vivante_fbdev/imx_display.c b/EXA/src/vivante_fbdev/imx_display.c
-index 876d37e..8d4eaa0 100644
---- a/EXA/src/vivante_fbdev/imx_display.c
-+++ b/EXA/src/vivante_fbdev/imx_display.c
-@@ -27,6 +27,8 @@
- 
- #include <fcntl.h>
- #include <errno.h>
-+#include <unistd.h>
-+#include <sys/ioctl.h>
-  
- #include <linux/fb.h>
- #include "xf86DDC.h"
-@@ -83,7 +85,7 @@ GCD(int a, int b)
- }
- 
- static int
--LCM(a, b)
-+LCM(int a, int b)
- {
- 	return (a * b) / GCD(a, b);
- }
-diff --git a/EXA/src/vivante_fbdev/vivante_fbdev_driver.c b/EXA/src/vivante_fbdev/vivante_fbdev_driver.c
-index ec407df..d35912a 100644
---- a/EXA/src/vivante_fbdev/vivante_fbdev_driver.c
-+++ b/EXA/src/vivante_fbdev/vivante_fbdev_driver.c
-@@ -29,8 +29,10 @@
- #include "vivante.h"
- #include "vivante_exa.h"
- #include "vivante_ext.h"
-+#include "vivante_dri.h"
- #include "imx_display.h"
- #include <errno.h>
-+#include <sys/ioctl.h>
- #include <linux/fb.h>
- #include <xorg/shmint.h>
- 
-diff --git a/EXA/src/vivante_gal/vivante_gal.h b/EXA/src/vivante_gal/vivante_gal.h
-index e186eda..332c773 100644
---- a/EXA/src/vivante_gal/vivante_gal.h
-+++ b/EXA/src/vivante_gal/vivante_gal.h
-@@ -260,6 +260,7 @@ extern "C" {
-     Bool VIV2DCacheOperation(GALINFOPTR galInfo, Viv2DPixmapPtr ppix, VIVFLUSHTYPE flush_type, int onLinearBuffer);
-     Bool VIV2DGPUUserMemMap(char* logical, unsigned int physical, unsigned int size, void ** mappingInfo, unsigned int * gpuAddress);
-     Bool VIV2DGPUUserMemUnMap(char* logical, unsigned int size, void * mappingInfo, unsigned int gpuAddress);
-+    Bool VIV2DGPUSurfaceReAllocNonCached(GALINFOPTR galInfo, Viv2DPixmapPtr ppriv);
-     Bool MapUserMemToGPU(GALINFOPTR galInfo, MemMapInfoPtr mmInfo);
-     void UnmapUserMem(GALINFOPTR galInfo, MemMapInfoPtr mmInfo);
-     /************************************************************************
-diff --git a/EXA/src/vivante_gal/vivante_gal_surface.c b/EXA/src/vivante_gal/vivante_gal_surface.c
-old mode 100755
-new mode 100644
-index 036b61b..0234814
---- a/EXA/src/vivante_gal/vivante_gal_surface.c
-+++ b/EXA/src/vivante_gal/vivante_gal_surface.c
-@@ -673,7 +673,8 @@ static gctBOOL VIV2DGPUSurfaceAlloc(VIVGPUPtr gpuctx, gctUINT alignedWidth, gctU
-     return VIV2DGPUSurfaceAllocEx(gpuctx, alignedWidth, alignedHeight, bytesPerPixel, surface, getPixmapCachePolicy());
- }
- 
--gctBOOL VIV2DGPUSurfaceReAllocNonCached(VIVGPUPtr gpuctx, Viv2DPixmapPtr ppriv) {
-+Bool VIV2DGPUSurfaceReAllocNonCached(GALINFOPTR galInfo, Viv2DPixmapPtr ppriv) {
-+    VIVGPUPtr gpuctx = (VIVGPUPtr)galInfo->mGpu;
-     GenericSurfacePtr oldSurf = gcvNULL;
-     GenericSurfacePtr newSurf = gcvNULL;
-     gctUINT32 alignedWidth;
-diff --git a/EXA/src/vivante_gal/vivante_priv.h b/EXA/src/vivante_gal/vivante_priv.h
-index e8ee3d2..05f5748 100644
---- a/EXA/src/vivante_gal/vivante_priv.h
-+++ b/EXA/src/vivante_gal/vivante_priv.h
-@@ -29,6 +29,7 @@ extern "C" {
- #include "HAL/gc_hal.h"
- #include "HAL/gc_hal_raster.h"
- #include "HAL/gc_hal_base.h"
-+#include "vivante_debug.h"
- 
-     /************************************************************************
-      * PIXMAP_HANDLING_STUFF(START)
-@@ -97,6 +98,9 @@ extern "C" {
-      * DRIVER & DEVICE  Structs (END)
-      *************************************************************************/
- 
-+    gctBOOL VIV2DGPUSurfaceAllocEx(VIVGPUPtr gpuctx, gctUINT alignedWidth, gctUINT alignedHeight,
-+        gctUINT bytesPerPixel, GenericSurfacePtr * surface, enum PixmapCachePolicy cachePolicy);
-+
- #ifdef __cplusplus
- }
- #endif
--- 
-1.8.4.5
-
diff --git a/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante_5.0.11.p7.1.bb b/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante_5.0.11.p7.4.bb
similarity index 48%
rename from recipes-graphics/xorg-driver/xf86-video-imxfb-vivante_5.0.11.p7.1.bb
rename to recipes-graphics/xorg-driver/xf86-video-imxfb-vivante_5.0.11.p7.4.bb
index 9fdd73a..4c8af46 100644
--- a/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante_5.0.11.p7.1.bb
+++ b/recipes-graphics/xorg-driver/xf86-video-imxfb-vivante_5.0.11.p7.4.bb
@@ -4,9 +4,8 @@
 
 require xf86-video-imxfb-vivante.inc
 
-SRC_URI += "file://0001-MGS-907-X11-EXA-Warnings-causes-GCC5.2-build-to-fail.patch \
-            file://Stop-using-Git-to-write-local-version.patch \
+SRC_URI += "file://Stop-using-Git-to-write-local-version.patch \
             file://Remove-dix-internal-header-usage.patch"
 
-SRC_URI[md5sum] = "05c54ed6d63e2a21dc18219e70869c29"
-SRC_URI[sha256sum] = "ab6a896cafba324105d26f52b7dad2c41d04d75cc9e48b66e8041a95660a04ee"
+SRC_URI[md5sum] = "6e2d8ee518e9eef8421e9dcdc1ea79c3"
+SRC_URI[sha256sum] = "3effaf42bde64da8fad2dfe0dcef12d1cfbbbe189511dbd4612442129fd832cc"
-- 
1.9.1

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Richard Röjfors]

Hi Neena,

I tried this driver with weston.

When running with two framebuffers like this:
/usr/bin/weston --tty=1 --use-gal2d=1 --use-gl=0 --device=/dev/fb2,/dev/fb0


I get a crash in the second call to gcoOS_InitLocalDisplayInfo in
gal2d_renderer_output_create (when the second output is to be setup).
The interesting thing is that its the same display  passed in as in
the first call, which succeeds...

Does this work on your end?

Thanks in advance,
Richard

On Tue, Jan 5, 2016 at 3:39 AM, Neena Busireddy
<neenareddy.busireddy@nxp.com> wrote:
> From: Neena Busireddy <neena.busireddy@freescale.com>
>
> This version is released with 3.14.52-1.1.0_ga release.
>
> Includes many of the bug fixes and stability improvements.
> For more information refer to i.MX Linux Release Notes from following link:
> http://www.nxp.com/products/microcontrollers-and-processors/arm-processors/i.mx-applications-processors-based-on-arm-cores/i.mx-6-processors/i.mx6qp/i.mx-6quad-processors-high-performance-3d-graphics-hd-video-arm-cortex-a9-core:i.MX6Q?fpsp=1&tab=Documentation_Tab#nogo
>
> L3.14.52_1.1.0_LINUX_DOCS package is under Supporting Information.
>
> Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
> Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
> ---
>  ...{imx-gpu-viv_5.0.11.p7.1-hfp.bb => imx-gpu-viv_5.0.11.p7.4-hfp.bb} | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>  rename recipes-graphics/imx-gpu-viv/{imx-gpu-viv_5.0.11.p7.1-hfp.bb => imx-gpu-viv_5.0.11.p7.4-hfp.bb} (54%)
>
> diff --git a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
> similarity index 54%
> rename from recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
> rename to recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
> index 22f097b..7895b6a 100644
> --- a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
> +++ b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
> @@ -3,7 +3,7 @@
>
>  require imx-gpu-viv.inc
>
> -SRC_URI[md5sum] = "b26211f322c325e76f912f432989f2ee"
> -SRC_URI[sha256sum] = "e49ca1d31ca83bb50f6672325db37ae3754eb9073ad98a37790ab1209ea5cba8"
> +SRC_URI[md5sum] = "d348def4c050e215efe970828752e5b9"
> +SRC_URI[sha256sum] = "b80838104f2f2e3b0baf2f1aacdca263c1c92589ce919b3be40aaad23f686b7f"
>
>  PACKAGE_FP_TYPE = "hardfp"
> --
> 1.9.1
>
> --
> _______________________________________________
> meta-freescale mailing list
> meta-freescale@yoctoproject.org
> https://lists.yoctoproject.org/listinfo/meta-freescale

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Richard Röjfors]

Hi again,

I also hit a deadlock on exit when running on one framebuffer like
this: (and when pressing ctrl-c it hangs forever)
/usr/bin/weston --tty=1 --use-gal2d=1 --use-gl=0 --device=/dev/fb2

The worker thread and cleanup code didn't look tight so I changed it a
bit and avoid the deadlock.
Will you fix this our should I send a patch?

Here is what I did:

--- weston-1.8.0/src/gal2d-renderer.c.org 2016-01-05 15:08:48.000000000 +0100
+++ weston-1.8.0/src/gal2d-renderer.c 2016-01-05 15:12:41.000000000 +0100
@@ -495,21 +495,21 @@
 {
     struct weston_output *output = (struct weston_output *)arg;
     struct gal2d_output_state *go = get_output_state(output);
+    int go_on = 1;

-    while(1)
+    while(go_on)
     {
-        if(gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE) == gcvSTATUS_OK )
+        int ok = gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE)
== gcvSTATUS_OK;
+
+        pthread_mutex_lock(&go->workerMutex);
+        go_on = go->exitWorker == 0;
+        pthread_mutex_unlock(&go->workerMutex);
+
+        if (ok && go_on)
         {
             gal2d_flip_surface(output);
             gcoOS_Signal(gcvNULL,go->busySignal, gcvTRUE);
         }
-        pthread_mutex_lock(&go->workerMutex);
-        if(go->exitWorker == 1)
-        {
-            pthread_mutex_unlock(&go->workerMutex);
-            break;
-        }
-        pthread_mutex_unlock(&go->workerMutex);
     }
     return 0;
 }
@@ -1169,10 +1169,10 @@
  }
     else
     {
-        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
         pthread_mutex_lock(&go->workerMutex);
         go->exitWorker = 1;
         pthread_mutex_unlock(&go->workerMutex);
+        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
         pthread_join(go->workerId, NULL);
     }

On Tue, Jan 5, 2016 at 12:42 PM, Richard Röjfors
<richard.rojfors@gmail.com> wrote:
> Hi Neena,
>
> I tried this driver with weston.
>
> When running with two framebuffers like this:
> /usr/bin/weston --tty=1 --use-gal2d=1 --use-gl=0 --device=/dev/fb2,/dev/fb0
>
>
> I get a crash in the second call to gcoOS_InitLocalDisplayInfo in
> gal2d_renderer_output_create (when the second output is to be setup).
> The interesting thing is that its the same display  passed in as in
> the first call, which succeeds...
>
> Does this work on your end?
>
> Thanks in advance,
> Richard
>
> On Tue, Jan 5, 2016 at 3:39 AM, Neena Busireddy
> <neenareddy.busireddy@nxp.com> wrote:
>> From: Neena Busireddy <neena.busireddy@freescale.com>
>>
>> This version is released with 3.14.52-1.1.0_ga release.
>>
>> Includes many of the bug fixes and stability improvements.
>> For more information refer to i.MX Linux Release Notes from following link:
>> http://www.nxp.com/products/microcontrollers-and-processors/arm-processors/i.mx-applications-processors-based-on-arm-cores/i.mx-6-processors/i.mx6qp/i.mx-6quad-processors-high-performance-3d-graphics-hd-video-arm-cortex-a9-core:i.MX6Q?fpsp=1&tab=Documentation_Tab#nogo
>>
>> L3.14.52_1.1.0_LINUX_DOCS package is under Supporting Information.
>>
>> Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
>> Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
>> ---
>>  ...{imx-gpu-viv_5.0.11.p7.1-hfp.bb => imx-gpu-viv_5.0.11.p7.4-hfp.bb} | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>  rename recipes-graphics/imx-gpu-viv/{imx-gpu-viv_5.0.11.p7.1-hfp.bb => imx-gpu-viv_5.0.11.p7.4-hfp.bb} (54%)
>>
>> diff --git a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
>> similarity index 54%
>> rename from recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
>> rename to recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
>> index 22f097b..7895b6a 100644
>> --- a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
>> +++ b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
>> @@ -3,7 +3,7 @@
>>
>>  require imx-gpu-viv.inc
>>
>> -SRC_URI[md5sum] = "b26211f322c325e76f912f432989f2ee"
>> -SRC_URI[sha256sum] = "e49ca1d31ca83bb50f6672325db37ae3754eb9073ad98a37790ab1209ea5cba8"
>> +SRC_URI[md5sum] = "d348def4c050e215efe970828752e5b9"
>> +SRC_URI[sha256sum] = "b80838104f2f2e3b0baf2f1aacdca263c1c92589ce919b3be40aaad23f686b7f"
>>
>>  PACKAGE_FP_TYPE = "hardfp"
>> --
>> 1.9.1
>>
>> --
>> _______________________________________________
>> meta-freescale mailing list
>> meta-freescale@yoctoproject.org
>> https://lists.yoctoproject.org/listinfo/meta-freescale

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Neena Busireddy]

Hi Richard,

Hi again,

I also hit a deadlock on exit when running on one framebuffer like
this: (and when pressing ctrl-c it hangs forever) /usr/bin/weston --tty=1 --use-gal2d=1 --use-gl=0 --device=/dev/fb2

The worker thread and cleanup code didn't look tight so I changed it a bit and avoid the deadlock.
Will you fix this our should I send a patch?

Here is what I did:

--- weston-1.8.0/src/gal2d-renderer.c.org 2016-01-05 15:08:48.000000000 +0100
+++ weston-1.8.0/src/gal2d-renderer.c 2016-01-05 15:12:41.000000000 
+++ +0100
@@ -495,21 +495,21 @@
 {
     struct weston_output *output = (struct weston_output *)arg;
     struct gal2d_output_state *go = get_output_state(output);
+    int go_on = 1;

-    while(1)
+    while(go_on)
     {
-        if(gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE) == gcvSTATUS_OK )
+        int ok = gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE)
== gcvSTATUS_OK;
+
+        pthread_mutex_lock(&go->workerMutex);
+        go_on = go->exitWorker == 0;
+        pthread_mutex_unlock(&go->workerMutex);
+
+        if (ok && go_on)
         {
             gal2d_flip_surface(output);
             gcoOS_Signal(gcvNULL,go->busySignal, gcvTRUE);
         }
-        pthread_mutex_lock(&go->workerMutex);
-        if(go->exitWorker == 1)
-        {
-            pthread_mutex_unlock(&go->workerMutex);
-            break;
-        }
-        pthread_mutex_unlock(&go->workerMutex);
     }
     return 0;
 }
@@ -1169,10 +1169,10 @@
  }
     else
     {
-        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
         pthread_mutex_lock(&go->workerMutex);
         go->exitWorker = 1;
         pthread_mutex_unlock(&go->workerMutex);
+        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
         pthread_join(go->workerId, NULL);
     }

Thanks for bringing this up. 

Hi Prabhu,

Could you please comment on this one?

Thanks,
Neena

On Tue, Jan 5, 2016 at 12:42 PM, Richard Röjfors <richard.rojfors@gmail.com> wrote:
> Hi Neena,
>
> I tried this driver with weston.
>
> When running with two framebuffers like this:
> /usr/bin/weston --tty=1 --use-gal2d=1 --use-gl=0 
> --device=/dev/fb2,/dev/fb0
>
>
> I get a crash in the second call to gcoOS_InitLocalDisplayInfo in 
> gal2d_renderer_output_create (when the second output is to be setup).
> The interesting thing is that its the same display  passed in as in 
> the first call, which succeeds...
>
> Does this work on your end?
>
> Thanks in advance,
> Richard
>
> On Tue, Jan 5, 2016 at 3:39 AM, Neena Busireddy 
> <neenareddy.busireddy@nxp.com> wrote:
>> From: Neena Busireddy <neena.busireddy@freescale.com>
>>
>> This version is released with 3.14.52-1.1.0_ga release.
>>
>> Includes many of the bug fixes and stability improvements.
>> For more information refer to i.MX Linux Release Notes from following link:
>> http://www.nxp.com/products/microcontrollers-and-processors/arm-proce
>> ssors/i.mx-applications-processors-based-on-arm-cores/i.mx-6-processo
>> rs/i.mx6qp/i.mx-6quad-processors-high-performance-3d-graphics-hd-vide
>> o-arm-cortex-a9-core:i.MX6Q?fpsp=1&tab=Documentation_Tab#nogo
>>
>> L3.14.52_1.1.0_LINUX_DOCS package is under Supporting Information.
>>
>> Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com>
>> Signed-off-by: Neena Busireddy <neenareddy.busireddy@nxp.com>
>> ---
>>  ...{imx-gpu-viv_5.0.11.p7.1-hfp.bb => 
>> imx-gpu-viv_5.0.11.p7.4-hfp.bb} | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)  rename 
>> recipes-graphics/imx-gpu-viv/{imx-gpu-viv_5.0.11.p7.1-hfp.bb => 
>> imx-gpu-viv_5.0.11.p7.4-hfp.bb} (54%)
>>
>> diff --git 
>> a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb 
>> b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
>> similarity index 54%
>> rename from 
>> recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
>> rename to recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
>> index 22f097b..7895b6a 100644
>> --- a/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.1-hfp.bb
>> +++ b/recipes-graphics/imx-gpu-viv/imx-gpu-viv_5.0.11.p7.4-hfp.bb
>> @@ -3,7 +3,7 @@
>>
>>  require imx-gpu-viv.inc
>>
>> -SRC_URI[md5sum] = "b26211f322c325e76f912f432989f2ee"
>> -SRC_URI[sha256sum] = "e49ca1d31ca83bb50f6672325db37ae3754eb9073ad98a37790ab1209ea5cba8"
>> +SRC_URI[md5sum] = "d348def4c050e215efe970828752e5b9"
>> +SRC_URI[sha256sum] = "b80838104f2f2e3b0baf2f1aacdca263c1c92589ce919b3be40aaad23f686b7f"
>>
>>  PACKAGE_FP_TYPE = "hardfp"
>> --
>> 1.9.1
>>
>> --
>> _______________________________________________
>> meta-freescale mailing list
>> meta-freescale@yoctoproject.org
>> https://lists.yoctoproject.org/listinfo/meta-freescale

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Otavio Salvador]

On Tue, Jan 5, 2016 at 1:03 PM, Neena Busireddy
<neenareddy.busireddy@nxp.com> wrote:
> Hi Richard,
>
> Hi again,
>
> I also hit a deadlock on exit when running on one framebuffer like
> this: (and when pressing ctrl-c it hangs forever) /usr/bin/weston --tty=1 --use-gal2d=1 --use-gl=0 --device=/dev/fb2
>
> The worker thread and cleanup code didn't look tight so I changed it a bit and avoid the deadlock.
> Will you fix this our should I send a patch?
>
> Here is what I did:
>
> --- weston-1.8.0/src/gal2d-renderer.c.org 2016-01-05 15:08:48.000000000 +0100
> +++ weston-1.8.0/src/gal2d-renderer.c 2016-01-05 15:12:41.000000000
> +++ +0100
> @@ -495,21 +495,21 @@
>  {
>      struct weston_output *output = (struct weston_output *)arg;
>      struct gal2d_output_state *go = get_output_state(output);
> +    int go_on = 1;
>
> -    while(1)
> +    while(go_on)
>      {
> -        if(gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE) == gcvSTATUS_OK )
> +        int ok = gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE)
> == gcvSTATUS_OK;
> +
> +        pthread_mutex_lock(&go->workerMutex);
> +        go_on = go->exitWorker == 0;
> +        pthread_mutex_unlock(&go->workerMutex);
> +
> +        if (ok && go_on)
>          {
>              gal2d_flip_surface(output);
>              gcoOS_Signal(gcvNULL,go->busySignal, gcvTRUE);
>          }
> -        pthread_mutex_lock(&go->workerMutex);
> -        if(go->exitWorker == 1)
> -        {
> -            pthread_mutex_unlock(&go->workerMutex);
> -            break;
> -        }
> -        pthread_mutex_unlock(&go->workerMutex);
>      }
>      return 0;
>  }
> @@ -1169,10 +1169,10 @@
>   }
>      else
>      {
> -        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
>          pthread_mutex_lock(&go->workerMutex);
>          go->exitWorker = 1;
>          pthread_mutex_unlock(&go->workerMutex);
> +        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
>          pthread_join(go->workerId, NULL);
>      }
>
> Thanks for bringing this up.
>
> Hi Prabhu,
>
> Could you please comment on this one?

Prabhu, please comment.

Richard, please send a formal patch on top of master-next (I applied
the patches there for test).

-- 
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

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Prabhu Sundararaj]

We have the patch but unfortunately not included to 5.0.11P7.4. so we provided a GPU driver binary patch to fix the issue as workaround.

Best Regards,
Prabhu

-----Original Message-----
From: Otavio Salvador [mailto:otavio.salvador@ossystems.com.br] 
Sent: Monday, January 11, 2016 10:19 AM
To: Neena Busireddy <neenareddy.busireddy@nxp.com>
Cc: Richard Röjfors <richard.rojfors@gmail.com>; Prabhu Sundararaj <prabhu.sundararaj@nxp.com>; meta-freescale@yoctoproject.org
Subject: Re: [meta-freescale] [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

On Tue, Jan 5, 2016 at 1:03 PM, Neena Busireddy <neenareddy.busireddy@nxp.com> wrote:
> Hi Richard,
>
> Hi again,
>
> I also hit a deadlock on exit when running on one framebuffer like
> this: (and when pressing ctrl-c it hangs forever) /usr/bin/weston 
> --tty=1 --use-gal2d=1 --use-gl=0 --device=/dev/fb2
>
> The worker thread and cleanup code didn't look tight so I changed it a bit and avoid the deadlock.
> Will you fix this our should I send a patch?
>
> Here is what I did:
>
> --- weston-1.8.0/src/gal2d-renderer.c.org 2016-01-05 
> 15:08:48.000000000 +0100
> +++ weston-1.8.0/src/gal2d-renderer.c 2016-01-05 15:12:41.000000000
> +++ +0100
> @@ -495,21 +495,21 @@
>  {
>      struct weston_output *output = (struct weston_output *)arg;
>      struct gal2d_output_state *go = get_output_state(output);
> +    int go_on = 1;
>
> -    while(1)
> +    while(go_on)
>      {
> -        if(gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE) == gcvSTATUS_OK )
> +        int ok = gcoOS_WaitSignal(gcvNULL, go->signal, gcvINFINITE)
> == gcvSTATUS_OK;
> +
> +        pthread_mutex_lock(&go->workerMutex);
> +        go_on = go->exitWorker == 0;
> +        pthread_mutex_unlock(&go->workerMutex);
> +
> +        if (ok && go_on)
>          {
>              gal2d_flip_surface(output);
>              gcoOS_Signal(gcvNULL,go->busySignal, gcvTRUE);
>          }
> -        pthread_mutex_lock(&go->workerMutex);
> -        if(go->exitWorker == 1)
> -        {
> -            pthread_mutex_unlock(&go->workerMutex);
> -            break;
> -        }
> -        pthread_mutex_unlock(&go->workerMutex);
>      }
>      return 0;
>  }
> @@ -1169,10 +1169,10 @@
>   }
>      else
>      {
> -        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
>          pthread_mutex_lock(&go->workerMutex);
>          go->exitWorker = 1;
>          pthread_mutex_unlock(&go->workerMutex);
> +        gcoOS_Signal(gcvNULL,go->signal, gcvTRUE);
>          pthread_join(go->workerId, NULL);
>      }
>
> Thanks for bringing this up.
>
> Hi Prabhu,
>
> Could you please comment on this one?

Prabhu, please comment.

Richard, please send a formal patch on top of master-next (I applied the patches there for test).

-- 
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

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Otavio Salvador]

On Mon, Jan 11, 2016 at 2:55 PM, Prabhu Sundararaj
<prabhu.sundararaj@nxp.com> wrote:
> We have the patch but unfortunately not included to 5.0.11P7.4. so we provided a GPU driver binary patch to fix the issue as workaround.

Sorry but it does not make sense for me.

The patch here seems simple. What binary update are you talking about?

-- 
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

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Prabhu Sundararaj]

libGAL.so need to be updated which is part of imx-gpu-viv.5.0.11.p7.4.bin

-----Original Message-----
From: Otavio Salvador [mailto:otavio.salvador@ossystems.com.br] 
Sent: Monday, January 11, 2016 11:30 AM
To: Prabhu Sundararaj <prabhu.sundararaj@nxp.com>
Cc: Neena Busireddy <neenareddy.busireddy@nxp.com>; Richard Röjfors <richard.rojfors@gmail.com>; meta-freescale@yoctoproject.org
Subject: Re: [meta-freescale] [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

On Mon, Jan 11, 2016 at 2:55 PM, Prabhu Sundararaj <prabhu.sundararaj@nxp.com> wrote:
> We have the patch but unfortunately not included to 5.0.11P7.4. so we provided a GPU driver binary patch to fix the issue as workaround.

Sorry but it does not make sense for me.

The patch here seems simple. What binary update are you talking about?

-- 
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

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Otavio Salvador]

On Mon, Jan 11, 2016 at 4:32 PM, Prabhu Sundararaj
<prabhu.sundararaj@nxp.com> wrote:
> libGAL.so need to be updated which is part of imx-gpu-viv.5.0.11.p7.4.bin

So please make a new tarball release; we need to ensure the binaries
are provided by NXP and are valid.

-- 
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

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Prabhu Sundararaj]

Hi Otavio,
We identified the problem after 5.0.11P7.4 release. 
Fix for the issue is part of the GPU driver binary. We need to get through the QA to release the fix.
Planning to include the fix in the next release (1st quarter of 2016).

Best Regards,
Prabhu

-----Original Message-----
From: Prabhu Sundararaj 
Sent: Monday, January 11, 2016 12:32 PM
To: 'Otavio Salvador' <otavio.salvador@ossystems.com.br>
Cc: Neena Busireddy <neenareddy.busireddy@nxp.com>; Richard Röjfors <richard.rojfors@gmail.com>; meta-freescale@yoctoproject.org
Subject: RE: [meta-freescale] [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

libGAL.so need to be updated which is part of imx-gpu-viv.5.0.11.p7.4.bin

-----Original Message-----
From: Otavio Salvador [mailto:otavio.salvador@ossystems.com.br] 
Sent: Monday, January 11, 2016 11:30 AM
To: Prabhu Sundararaj <prabhu.sundararaj@nxp.com>
Cc: Neena Busireddy <neenareddy.busireddy@nxp.com>; Richard Röjfors <richard.rojfors@gmail.com>; meta-freescale@yoctoproject.org
Subject: Re: [meta-freescale] [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

On Mon, Jan 11, 2016 at 2:55 PM, Prabhu Sundararaj <prabhu.sundararaj@nxp.com> wrote:
> We have the patch but unfortunately not included to 5.0.11P7.4. so we provided a GPU driver binary patch to fix the issue as workaround.

Sorry but it does not make sense for me.

The patch here seems simple. What binary update are you talking about?

-- 
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

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Otavio Salvador]

On Mon, Jan 11, 2016 at 6:34 PM, Prabhu Sundararaj
<prabhu.sundararaj@nxp.com> wrote:
> We identified the problem after 5.0.11P7.4 release.
> Fix for the issue is part of the GPU driver binary. We need to get through the QA to release the fix.
> Planning to include the fix in the next release (1st quarter of 2016).

So Richard patch seems to be the best temporary fix; Richard, could
you cook a patch for it?

When new package is out, we revert the change.

-- 
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

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Richard Röjfors]

[-- Attachment #1: Type: text/plain, Size: 2377 bytes --]

Hi Both,

Just to clarify, we are talking about two different issues.

I think Octavio talks about my patch to solve the dead lock on exit, while
Prabhu talks about a patch needed to support multiple framebuffers.

"My" patch is on the open weston code (the gal2d-renderer), while the other
is in the NXP blobs.

I think both are of interest for the anyone on the list.

I'm happy to provide a proper patch for the dead lock. If you don't have
another solution Prabhu?

Thanks,
Richard

On Mon, Jan 11, 2016 at 9:34 PM, Prabhu Sundararaj <
prabhu.sundararaj@nxp.com> wrote:

> Hi Otavio,
> We identified the problem after 5.0.11P7.4 release.
> Fix for the issue is part of the GPU driver binary. We need to get through
> the QA to release the fix.
> Planning to include the fix in the next release (1st quarter of 2016).
>
> Best Regards,
> Prabhu
>
> -----Original Message-----
> From: Prabhu Sundararaj
> Sent: Monday, January 11, 2016 12:32 PM
> To: 'Otavio Salvador' <otavio.salvador@ossystems.com.br>
> Cc: Neena Busireddy <neenareddy.busireddy@nxp.com>; Richard Röjfors <
> richard.rojfors@gmail.com>; meta-freescale@yoctoproject.org
> Subject: RE: [meta-freescale] [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv:
> Update to 5.0.11-p7.4 version
>
> libGAL.so need to be updated which is part of imx-gpu-viv.5.0.11.p7.4.bin
>
> -----Original Message-----
> From: Otavio Salvador [mailto:otavio.salvador@ossystems.com.br]
> Sent: Monday, January 11, 2016 11:30 AM
> To: Prabhu Sundararaj <prabhu.sundararaj@nxp.com>
> Cc: Neena Busireddy <neenareddy.busireddy@nxp.com>; Richard Röjfors <
> richard.rojfors@gmail.com>; meta-freescale@yoctoproject.org
> Subject: Re: [meta-freescale] [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv:
> Update to 5.0.11-p7.4 version
>
> On Mon, Jan 11, 2016 at 2:55 PM, Prabhu Sundararaj <
> prabhu.sundararaj@nxp.com> wrote:
> > We have the patch but unfortunately not included to 5.0.11P7.4. so we
> provided a GPU driver binary patch to fix the issue as workaround.
>
> Sorry but it does not make sense for me.
>
> The patch here seems simple. What binary update are you talking about?
>
> --
> 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
>

[-- Attachment #2: Type: text/html, Size: 3979 bytes --]

Re: [meta-fsl-arm][PATCH v3 1/5] imx-gpu-viv: Update to 5.0.11-p7.4 version

[Otavio Salvador]

On Mon, Jan 11, 2016 at 7:01 PM, Richard Röjfors
<richard.rojfors@gmail.com> wrote:
> Just to clarify, we are talking about two different issues.
>
> I think Octavio talks about my patch to solve the dead lock on exit, while
> Prabhu talks about a patch needed to support multiple framebuffers.
>
> "My" patch is on the open weston code (the gal2d-renderer), while the other
> is in the NXP blobs.
>
> I think both are of interest for the anyone on the list.
>
> I'm happy to provide a proper patch for the dead lock. If you don't have
> another solution Prabhu?

The patch, please prepare it. It is important indeed.

I think the binary fix need to be formally released so we can upgrade
the packages and have it incorporated. There is no justification to
drop a binary .so file without proper release as it removes all
support responsibility and is clearly wrong.

Anyone using a patched binary will be using non-supported software; so
a formal release of the binary should be done.

-- 
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