Re: [oe] [oe-commits] Armin Kuster : netmap: add new package

[ChenQi]

Hi Armin,

I found that if CONFIG_IXGBE is enabled in kernel. This package fails to 
build with error messages like below.


|  #error "unsupported ixgbe driver version"
|   ^

...
...

error: request for member 'count' in something not a structure or union
|   na.num_rx_desc = NM_IXGBE_RX_RING(adapter, 0)->count;
|^
| cc1: some warnings being treated as errors


Do you have any idea how to fix it? Thanks.

Best Regards,
Chen Qi

On 09/17/2015 11:04 PM, g...@opal.openembedded.org wrote:

Module: meta-openembedded.git
Branch: master-next
Commit: 1692d5c3020434404fc1ee6911a60b88287a5efb
URL:
http://git.openembedded.org/?p=meta-openembedded.git&a=commit;h=1692d5c3020434404fc1ee6911a60b88287a5efb

Author: Armin Kuster 
Date:   Sun Aug  9 05:56:30 2015 +0530

netmap: add new package

this adds netmap, the fast packet I/O framework
http://info.iet.unipi.it/~luigi/netmap/

 - updated to latest version to get kernel 4.1 support
 - fixed printf type issue
 - Fixed manual config options

Signed-off-by: Armin Kuster 
Signed-off-by: Martin Jansa 
Signed-off-by: Joe MacDonald 

---

  ...1-testmmap-fix-compile-issue-with-gcc-5.x.patch | 33 
  .../netmap/files/makefile_fixup.patch  | 97 ++
  .../recipes-kernel/netmap/netmap-modules_git.bb| 57 +
  meta-networking/recipes-kernel/netmap/netmap.inc   | 17 
  .../recipes-kernel/netmap/netmap_git.bb| 35 
  5 files changed, 239 insertions(+)

diff --git 
a/meta-networking/recipes-kernel/netmap/files/0001-testmmap-fix-compile-issue-with-gcc-5.x.patch
 
b/meta-networking/recipes-kernel/netmap/files/0001-testmmap-fix-compile-issue-with-gcc-5.x.patch
new file mode 100644
index 000..ea36b1c
--- /dev/null
+++ 
b/meta-networking/recipes-kernel/netmap/files/0001-testmmap-fix-compile-issue-with-gcc-5.x.patch
@@ -0,0 +1,33 @@
+Upstream-Status: Pending
+
+From c81bf54d6eb870286662a11d3b4a994717c47696 Mon Sep 17 00:00:00 2001
+From: Armin Kuster 
+Date: Tue, 8 Sep 2015 05:36:27 -0700
+Subject: [PATCH] testmmap: fix compile issue with gcc 5.x
+
+this fixes:
+examples/testmmap.c:540:10: error: format '%ld' expects argument of type 'long 
int', but argument 3 has type 'ssize_t {aka const int}' [-Werror=format=]
+|printf("ring_ofs[%d] %ld\n", i, nifp->ring_ofs[i]);
+|   ^
+
+Signed-off-by: Armin Kuster 
+---
+ examples/testmmap.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/examples/testmmap.c b/examples/testmmap.c
+index d7f6acc..934489a 100644
+--- a/examples/testmmap.c
 b/examples/testmmap.c
+@@ -537,7 +537,7 @@ do_if()
+   for (i = 0; i < 5; i++)
+   printf("spare1[%d]  %u\n", i, nifp->ni_spare1[i]);
+   for (i = 0; i < (nifp->ni_tx_rings + nifp->ni_rx_rings + 2); i++)
+-  printf("ring_ofs[%d] %ld\n", i, nifp->ring_ofs[i]);
++  printf("ring_ofs[%d] %zd\n", i, nifp->ring_ofs[i]);
+ }
+
+ struct netmap_ring *
+--
+2.3.5
+
diff --git a/meta-networking/recipes-kernel/netmap/files/makefile_fixup.patch 
b/meta-networking/recipes-kernel/netmap/files/makefile_fixup.patch
new file mode 100644
index 000..b050bf0
--- /dev/null
+++ b/meta-networking/recipes-kernel/netmap/files/makefile_fixup.patch
@@ -0,0 +1,97 @@
+The makefile assumes building locally.
+
+Upstream Status: Inappropriate [Native]
+
+The configure is not auto-make based
+
+Signed-of-by: Armin Kuster 
+
+Index: LINUX/netmap.mak.in
+===
+--- a/LINUX/netmap.mak.in
 b/LINUX/netmap.mak.in
+@@ -12,10 +12,9 @@ SRCDIR:=@SRCDIR@
+ # The following commands are needed to build the modules as out-of-tree,
+ # in fact the kernel sources path must be specified.
+
+-PWD ?= $(CURDIR)
+
+ # Additional compile flags (e.g. header location)
+-EXTRA_CFLAGS := -I$(PWD) -I$(SRCDIR) -I$(SRCDIR)/../sys 
-I$(SRCDIR)/../sys/dev -DCONFIG_NETMAP
++EXTRA_CFLAGS := -I$(SRCDIR) -I$(SRCDIR)/../sys -I$(SRCDIR)/../sys/dev 
-DCONFIG_NETMAP
+ EXTRA_CFLAGS += -Wno-unused-but-set-variable
+ EXTRA_CFLAGS += $(foreach s,$(SUBSYS),-DCONFIG_NETMAP_$(shell echo $s|tr a-z- 
A-Z_))
+
+Index: LINUX/configure
+===
+--- a/LINUX/configure
 b/LINUX/configure
+@@ -349,52 +349,6 @@ reset_tests() {
+   EOF
+ }
+
+-# run_tests: run all accumulated tests and exec the pertinent
+-#   success/failure actions for each one.
+-run_tests() {
+-  local t=# prevent -EOF to eat the make TAB
+-  cat > $TMPDIR/Makefile <<-EOF
+-  ifneq (\$(KERNELRELEASE),)
+-  obj-m := $TESTOBJS
+-  else
+-  S_DRIVERS := $(drv print)
+-  all: \$(S_DRIVERS:%=get-%)
+-  $t  \$(MAKE) -C $ksrc M=\$\$PWD $kopts
+-
+-  -include $TOPDIR/drivers.mak
+-  EOF
+-  for d in $(drv print); do
+-  cat >> $TMPDIR/Makefile <<-EOF
+- 

Re: [oe] [oe-commits] Armin Kuster : netmap: add new package

[ChenQi]

On 09/30/2015 12:03 AM, akuster wrote:

Chen Qi,

  It looks like "NETMAP_LINUX_IXGBE_DESC" needs to be defined. Supported
values are 1-3.

That variable can be appended to ${S}/LINUX/netmap_linux_config.h. Maybe
it needs to be added to the recipe?


I tried these three values. None of them work.



if not try add to the EXTRA_OECONF:

--no-drivers=   do not compile the given drivers (comma sep.)
or
--drivers=  only compile the given drivers (comma sep.)


EXTRA_OECONF += "--no-drivers=ixgbe"
The above configuration works.

Best Regards,
Chen Qi


Let me know what works.

regards,
Armin


On 09/28/2015 11:00 PM, ChenQi wrote:

Hi Armin,

I found that if CONFIG_IXGBE is enabled in kernel. This package fails to
build with error messages like below.

|  #error "unsupported ixgbe driver version"
|   ^

...
...

error: request for member 'count' in something not a structure or union
|   na.num_rx_desc = NM_IXGBE_RX_RING(adapter, 0)->count;
|^
| cc1: some warnings being treated as errors


Do you have any idea how to fix it? Thanks.

Best Regards,
Chen Qi

On 09/17/2015 11:04 PM, g...@opal.openembedded.org wrote:

Module: meta-openembedded.git
Branch: master-next
Commit: 1692d5c3020434404fc1ee6911a60b88287a5efb
URL:
http://git.openembedded.org/?p=meta-openembedded.git&a=commit;h=1692d5c3020434404fc1ee6911a60b88287a5efb


Author: Armin Kuster 
Date:   Sun Aug  9 05:56:30 2015 +0530

netmap: add new package

this adds netmap, the fast packet I/O framework
http://info.iet.unipi.it/~luigi/netmap/

  - updated to latest version to get kernel 4.1 support
  - fixed printf type issue
  - Fixed manual config options

Signed-off-by: Armin Kuster 
Signed-off-by: Martin Jansa 
Signed-off-by: Joe MacDonald 

---

   ...1-testmmap-fix-compile-issue-with-gcc-5.x.patch | 33 
   .../netmap/files/makefile_fixup.patch  | 97
++
   .../recipes-kernel/netmap/netmap-modules_git.bb| 57 +
   meta-networking/recipes-kernel/netmap/netmap.inc   | 17 
   .../recipes-kernel/netmap/netmap_git.bb| 35 
   5 files changed, 239 insertions(+)

diff --git
a/meta-networking/recipes-kernel/netmap/files/0001-testmmap-fix-compile-issue-with-gcc-5.x.patch
b/meta-networking/recipes-kernel/netmap/files/0001-testmmap-fix-compile-issue-with-gcc-5.x.patch

new file mode 100644
index 000..ea36b1c
--- /dev/null
+++
b/meta-networking/recipes-kernel/netmap/files/0001-testmmap-fix-compile-issue-with-gcc-5.x.patch

@@ -0,0 +1,33 @@
+Upstream-Status: Pending
+
+From c81bf54d6eb870286662a11d3b4a994717c47696 Mon Sep 17 00:00:00 2001
+From: Armin Kuster 
+Date: Tue, 8 Sep 2015 05:36:27 -0700
+Subject: [PATCH] testmmap: fix compile issue with gcc 5.x
+
+this fixes:
+examples/testmmap.c:540:10: error: format '%ld' expects argument of
type 'long int', but argument 3 has type 'ssize_t {aka const int}'
[-Werror=format=]
+|printf("ring_ofs[%d] %ld\n", i, nifp->ring_ofs[i]);
+|   ^
+
+Signed-off-by: Armin Kuster 
+---
+ examples/testmmap.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/examples/testmmap.c b/examples/testmmap.c
+index d7f6acc..934489a 100644
+--- a/examples/testmmap.c
 b/examples/testmmap.c
+@@ -537,7 +537,7 @@ do_if()
+ for (i = 0; i < 5; i++)
+ printf("spare1[%d]  %u\n", i, nifp->ni_spare1[i]);
+ for (i = 0; i < (nifp->ni_tx_rings + nifp->ni_rx_rings + 2); i++)
+-printf("ring_ofs[%d] %ld\n", i, nifp->ring_ofs[i]);
++printf("ring_ofs[%d] %zd\n", i, nifp->ring_ofs[i]);
+ }
+
+ struct netmap_ring *
+--
+2.3.5
+
diff --git
a/meta-networking/recipes-kernel/netmap/files/makefile_fixup.patch
b/meta-networking/recipes-kernel/netmap/files/makefile_fixup.patch
new file mode 100644
index 000..b050bf0
--- /dev/null
+++ b/meta-networking/recipes-kernel/netmap/files/makefile_fixup.patch
@@ -0,0 +1,97 @@
+The makefile assumes building locally.
+
+Upstream Status: Inappropriate [Native]
+
+The configure is not auto-make based
+
+Signed-of-by: Armin Kuster 
+
+Index: LINUX/netmap.mak.in
+===
+--- a/LINUX/netmap.mak.in
 b/LINUX/netmap.mak.in
+@@ -12,10 +12,9 @@ SRCDIR:=@SRCDIR@
+ # The following commands are needed to build the modules as
out-of-tree,
+ # in fact the kernel sources path must be specified.
+
+-PWD ?= $(CURDIR)
+
+ # Additional compile flags (e.g. header location)
+-EXTRA_CFLAGS := -I$(PWD) -I$(SRCDIR) -I$(SRCDIR)/../sys
-I$(SRCDIR)/../sys/dev -DCONFIG_NETMAP
++EXTRA_CFLAGS := -I$(SRCDIR) -I$(SRCDIR)/../sys
-I$(SRCDIR)/../sys/dev -DCONFIG_NETMAP
+ EXTRA_CFLAGS += -Wno-unused-but-set-variable
+ EXTRA_CFLAGS += $(foreach s,$(SUBSYS),-DCONFIG_NETMAP_$(shell echo
$s|tr a-z- A-Z_))
+
+Index: LINUX/configure
+==

Re: [oe] [meta-networking][PATCH] netmap-modules: fix compilation failure

[ChenQi]

On 10/08/2015 04:33 PM, akuster808 wrote:

Chen Qi,


On 10/07/2015 10:02 PM, Chen Qi wrote:

If CONFIG_IXGBE is enabled in kernel. This package fails to build with
error messages like below.

|  #error "unsupported ixgbe driver version"
|   ^

...

error: request for member 'count' in something not a structure or union
|   na.num_rx_desc = NM_IXGBE_RX_RING(adapter, 0)->count;
|^
| cc1: some warnings being treated as errors

Fix this problem by disabling igxbe driver in netmap-modules.

Signed-off-by: Chen Qi 
---
  meta-networking/recipes-kernel/netmap/netmap-modules_git.bb | 2 ++
  1 file changed, 2 insertions(+)

diff --git a/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb 
b/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb
index f9ac9cc..c1c319d 100644
--- a/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb
+++ b/meta-networking/recipes-kernel/netmap/netmap-modules_git.bb
@@ -15,6 +15,8 @@ EXTRA_OECONF = "--kernel-dir=${STAGING_KERNEL_BUILDDIR} \
  --driver-suffix="-netmap" \
  "
  
+EXTRA_OECONF += "--no-drivers=ixgbe"

+

Would the following also work?

+PACKAGECONFIG ?= ""
+PACKAGECONFIG[IXGBE] = "--drivers=ixgbe,--no-drivers=ixgbe"

Then we replicate that for the other kernel modules.

- armin


The problem is that if 'IXGBE' is enabled, the compilation fails.
I tried three values for the macro, none of them works.

Best Regards,
Chen Qi


  LDFLAGS := "${@'${LDFLAGS}'.replace('-Wl,-O1', '')}"
  LDFLAGS := "${@'${LDFLAGS}'.replace('-Wl,--as-needed', '')}"
  





--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH 1/4] rrdtool: fix useless-rpath QA issue

[ChenQi]

On 10/20/2015 11:57 PM, Martin Jansa wrote:

On Tue, Oct 13, 2015 at 10:43:55AM +0800, Chen Qi wrote:

When building rrdtool, we would meet the following error.

Now it fails for qemuarm and qemux86, not sure which change is causing
that:

http://errors.yoctoproject.org/Errors/Details/20597/
http://errors.yoctoproject.org/Errors/Details/20601/



I'm looking into this problem today, although I currently have some 
problem reproducing it.

I'll look at the logs and try to find out something.

Regards,
Chen Qi


   ERROR: QA Issue: rrdtool-python: work/core2-64-poky-linux/rrdtool
   /1.5.4-r0/packages-split/rrdtool-python/usr/lib/python2.7/site-packages
   /rrdtool.so contains probably-redundant RPATH /usr/lib [useless-rpaths]

Fix this issue by removing the redundante RPATH section from rrdtool.so.

As we use '--disable-rpath' in EXTRA_OECONF, we don't need to the `sed'
line in do_configure.

Signed-off-by: Chen Qi 
---
  meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb | 6 +-
  1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb 
b/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
index 91a30cf..f2ded65 100644
--- a/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
+++ b/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
@@ -71,7 +71,6 @@ do_configure() {
  PYTHON_INCLUDES="-I=/usr/include/python${PYTHON_BASEVERSION}"' \
  ${S}/m4/acinclude.m4
  #remove the hardcoded $(libdir) rpath
-sed -i -e 's|--rpath=$(libdir)||g' ${S}/bindings/Makefile.am
  
  autotools_do_configure
  
@@ -100,6 +99,11 @@ do_configure() {

  ${B}/examples/*.pl
  }
  
+do_compile_append() {

+# Remove useless rpath
+find ${B}/bindings/python/build -name "rrdtool.so" | xargs chrpath -d
+}
+
  PACKAGES =+ "${PN}-perl ${PN}-python"
  
  FILES_${PN}-doc += "${datadir}/examples"

--
1.9.1



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH 1/4] rrdtool: fix useless-rpath QA issue

[ChenQi]

Hi Ray,

I just saw this email.
Thanks a lot for your suggestion.
I've sent out V2.

Best Regards,
Chen Qi


On 10/20/2015 07:14 PM, Ray, Ian (GE Healthcare) wrote:

Hi,


Signed-off-by: Chen Qi 
---
  meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb | 6 +-
  1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb b/meta-
oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
index 91a30cf..f2ded65 100644
--- a/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
+++ b/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
@@ -71,7 +71,6 @@ do_configure() {
  PYTHON_INCLUDES="-I=/usr/include/python${PYTHON_BASEVERSION}"'
\
  ${S}/m4/acinclude.m4
  #remove the hardcoded $(libdir) rpath
-sed -i -e 's|--rpath=$(libdir)||g' ${S}/bindings/Makefile.am

I tested locally a simpler patch:

@@ -71,7 +71,7 @@ do_configure() {
  PYTHON_INCLUDES="-I=/usr/include/python${PYTHON_BASEVERSION}"' \
  ${S}/m4/acinclude.m4
  #remove the hardcoded $(libdir) rpath
-sed -i -e 's|--rpath=$(libdir)||g' ${S}/bindings/Makefile.am
+sed -i -e 's|LD_RUN_PATH=$(libdir)||g' ${S}/bindings/Makefile.am



@@ -100,6 +99,11 @@ do_configure() {
  ${B}/examples/*.pl
  }

+do_compile_append() {
+# Remove useless rpath
+find ${B}/bindings/python/build -name "rrdtool.so" | xargs chrpath -d
+}

This fails for me since chrpath is not built for multilib.




--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH V3 1/4] rrdtool: fix useless-rpath QA issue

[ChenQi]

Please ignore these patches.
I'll send out patches based on the latest master.

// Qi

On 11/04/2015 03:36 PM, Chen Qi wrote:

When building rrdtool, we would meet the following error.

   ERROR: QA Issue: rrdtool-python: work/core2-64-poky-linux/rrdtool
   /1.5.4-r0/packages-split/rrdtool-python/usr/lib/python2.7/site-packages
   /rrdtool.so contains probably-redundant RPATH /usr/lib [useless-rpaths]

Fix this issue by removing 'LD_RUN_PATH=$(libdir)' when compiling rrdtool.so.

As we use '--disable-rpath' in EXTRA_OECONF, we don't need to the to remove
'--rpath=$(libdir)' in Makefile.am.

Signed-off-by: Chen Qi 
---
  ...1-bindings-Makefile.am-remove-LD_RUN_PATH.patch | 24 ++
  meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb  |  3 +--
  2 files changed, 25 insertions(+), 2 deletions(-)
  create mode 100644 
meta-oe/recipes-extended/rrdtool/files/0001-bindings-Makefile.am-remove-LD_RUN_PATH.patch

diff --git 
a/meta-oe/recipes-extended/rrdtool/files/0001-bindings-Makefile.am-remove-LD_RUN_PATH.patch
 
b/meta-oe/recipes-extended/rrdtool/files/0001-bindings-Makefile.am-remove-LD_RUN_PATH.patch
new file mode 100644
index 000..539caa5
--- /dev/null
+++ 
b/meta-oe/recipes-extended/rrdtool/files/0001-bindings-Makefile.am-remove-LD_RUN_PATH.patch
@@ -0,0 +1,24 @@
+Upstream-Status: Inappropriate [OE specific]
+
+Subject: bindings/Makefile.am: remove LD_RUN_PATH
+
+Signed-off-by: Chen Qi 
+---
+ bindings/Makefile.am | 1 -
+ 1 file changed, 1 deletion(-)
+
+diff --git a/bindings/Makefile.am b/bindings/Makefile.am
+index 54f5c23..8ea16b7 100644
+--- a/bindings/Makefile.am
 b/bindings/Makefile.am
+@@ -67,7 +67,6 @@ python:
+   && env \
+ABS_TOP_SRCDIR=${abs_top_srcdir} \
+ABS_TOP_BUILDDIR=${abs_top_builddir} \
+-   LD_RUN_PATH=$(libdir) \
+$(PYTHON) ${abs_srcdir}/$@/setup.py build_ext $(PYTHON_RPATH) \
+   && env \
+ABS_TOP_SRCDIR=${abs_top_srcdir} \
+--
+1.9.1
+
diff --git a/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb 
b/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
index 91a30cf..2cf2dab 100644
--- a/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
+++ b/meta-oe/recipes-extended/rrdtool/rrdtool_1.5.4.bb
@@ -11,6 +11,7 @@ PV = "1.5.4"
  
  SRC_URI = "\

  git://github.com/oetiker/rrdtool-1.x.git;branch=1.5 \
+file://0001-bindings-Makefile.am-remove-LD_RUN_PATH.patch \
  "
  
  S = "${WORKDIR}/git"

@@ -70,8 +71,6 @@ do_configure() {
  sed -i -e '/PYTHON_INCLUDES="-I${/c \
  PYTHON_INCLUDES="-I=/usr/include/python${PYTHON_BASEVERSION}"' \
  ${S}/m4/acinclude.m4
-#remove the hardcoded $(libdir) rpath
-sed -i -e 's|--rpath=$(libdir)||g' ${S}/bindings/Makefile.am
  
  autotools_do_configure
  


--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [RFT] Glibc 2.23 and binutils 2.26

[ChenQi]

On 01/17/2016 07:54 PM, Khem Raj wrote:

Hi all

upstream glibc and binutils release branches have been cut out and are being 
made ready for next release
I have put together update patchset for both of them

here are the branches

https://github.com/kraj/openembedded-core/tree/kraj/binutils-2.26
https://github.com/kraj/openembedded-core/tree/kraj/glibc-2.23

FYI These are still using autorev to ensure that we test tip of release branches

Please give them a shot in your environments and report any issues you 
encounter.

Thanks for help

-Khem







0001-Generate-relocatable-SDKs.patch is included in SRC_URI twice, which 
causes the following error when building binutils-crosssdk-XXX.


ERROR: Command Error: exit status: 1  Output:
Patch 0001-Generate-relocatable-SDKs.patch is already applied; check 
your series file

ERROR: Function failed: patch_do_patch

binutils-2.26.inc and binutils-crosssdk_2.26.bb both has this patch in 
its SRC_RUI.


Regards,
Chen Qi
--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [oe-commits] [meta-openembedded] 03/19: openconnect: add dependency to fix Makefile.am

[ChenQi]

Hi Martin,

I just realized that this patch is wrong. It actually creates circular 
dependency between main.c and version.c.

Please drop it.
Sorry.

Best Regards,
Chen Qi

On 05/11/2016 06:40 PM, g...@opal.openembedded.org wrote:

martin_jansa pushed a commit to branch master-next
in repository meta-openembedded.

commit efadf2fce5a3bbf4b7d6f8ee80f2fdfbadc86b06
Author: qi.c...@windriver.com 
AuthorDate: Fri May 6 17:02:17 2016 +0800

 openconnect: add dependency to fix Makefile.am
 
 main.c needs version.c which is generated at build time.

 Add this dependency to avoid the following error.
 
   ../git/main.c:78:21: fatal error: version.c: No such file or directory
 
 Signed-off-by: Chen Qi 

 Signed-off-by: Martin Jansa 
---
  .../0001-Makefile.am-add-missing-dependency.patch  | 28 ++
  .../openconnect/openconnect_git.bb |  3 ++-
  2 files changed, 30 insertions(+), 1 deletion(-)

diff --git 
a/meta-networking/recipes-connectivity/openconnect/files/0001-Makefile.am-add-missing-dependency.patch
 
b/meta-networking/recipes-connectivity/openconnect/files/0001-Makefile.am-add-missing-dependency.patch
new file mode 100644
index 000..c465bf1
--- /dev/null
+++ 
b/meta-networking/recipes-connectivity/openconnect/files/0001-Makefile.am-add-missing-dependency.patch
@@ -0,0 +1,28 @@
+Upstream-Status: Pending
+
+Subject: Makefile.am: add missing dependency
+
+Add missing dependency to avoid the following error.
+
+  ../git/main.c:78:21: fatal error: version.c: No such file or directory
+
+Signed-off-by: Chen Qi 
+---
+ Makefile.am | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/Makefile.am b/Makefile.am
+index d823e01..ea48ec1 100644
+--- a/Makefile.am
 b/Makefile.am
+@@ -112,6 +112,7 @@ EXTRA_DIST += $(shell cd "$(top_srcdir)" && \
+
+ DISTCLEANFILES = $(pkgconfig_DATA)
+
++main.c: version.c
+ main.o: version.c
+ version.c: $(library_srcs) $(lib_openssl_srcs) $(lib_gnutls_srcs) \
+  $(openconnect_SOURCES) Makefile.am configure.ac \
+--
+2.8.1
+
diff --git 
a/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb 
b/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb
index 6d3c252..d8fe439 100644
--- a/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb
+++ b/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb
@@ -8,7 +8,8 @@ RDEPENDS_${PN} = "vpnc"
  PV = "7.06"
  
  SRCREV = "35542d52202672b8c12ecc63867432128244013a"

-SRC_URI = "git://git.infradead.org/users/dwmw2/openconnect.git"
+SRC_URI = "git://git.infradead.org/users/dwmw2/openconnect.git \
+   file://0001-Makefile.am-add-missing-dependency.patch"
  
  S = "${WORKDIR}/git"
  



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [oe-commits] [meta-openembedded] 01/44: openconnect: add dependency to fix Makefile.am

[ChenQi]

Hi Martin,

Sorry, but please drop this patch. It's a wrong fix.

The below one is the right fix.
[oe-commits] [meta-openembedded] 02/44: openconnect: fix do_compile failure

Best Regards,
Chen Qi

On 07/11/2016 09:38 PM, g...@opal.openembedded.org wrote:

martin_jansa pushed a commit to branch master-next
in repository meta-openembedded.

commit 939c471c5ed2471bf457bf21dc219edde68d1a19
Author: qi.c...@windriver.com 
AuthorDate: Fri May 6 17:02:17 2016 +0800

 openconnect: add dependency to fix Makefile.am
 
 main.c needs version.c which is generated at build time.

 Add this dependency to avoid the following error.
 
   ../git/main.c:78:21: fatal error: version.c: No such file or directory
 
 Signed-off-by: Chen Qi 

 Signed-off-by: Martin Jansa 
---
  .../0001-Makefile.am-add-missing-dependency.patch  | 28 ++
  .../openconnect/openconnect_git.bb |  3 ++-
  2 files changed, 30 insertions(+), 1 deletion(-)

diff --git 
a/meta-networking/recipes-connectivity/openconnect/files/0001-Makefile.am-add-missing-dependency.patch
 
b/meta-networking/recipes-connectivity/openconnect/files/0001-Makefile.am-add-missing-dependency.patch
new file mode 100644
index 000..c465bf1
--- /dev/null
+++ 
b/meta-networking/recipes-connectivity/openconnect/files/0001-Makefile.am-add-missing-dependency.patch
@@ -0,0 +1,28 @@
+Upstream-Status: Pending
+
+Subject: Makefile.am: add missing dependency
+
+Add missing dependency to avoid the following error.
+
+  ../git/main.c:78:21: fatal error: version.c: No such file or directory
+
+Signed-off-by: Chen Qi 
+---
+ Makefile.am | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/Makefile.am b/Makefile.am
+index d823e01..ea48ec1 100644
+--- a/Makefile.am
 b/Makefile.am
+@@ -112,6 +112,7 @@ EXTRA_DIST += $(shell cd "$(top_srcdir)" && \
+
+ DISTCLEANFILES = $(pkgconfig_DATA)
+
++main.c: version.c
+ main.o: version.c
+ version.c: $(library_srcs) $(lib_openssl_srcs) $(lib_gnutls_srcs) \
+  $(openconnect_SOURCES) Makefile.am configure.ac \
+--
+2.8.1
+
diff --git 
a/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb 
b/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb
index 6d3c252..d8fe439 100644
--- a/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb
+++ b/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb
@@ -8,7 +8,8 @@ RDEPENDS_${PN} = "vpnc"
  PV = "7.06"
  
  SRCREV = "35542d52202672b8c12ecc63867432128244013a"

-SRC_URI = "git://git.infradead.org/users/dwmw2/openconnect.git"
+SRC_URI = "git://git.infradead.org/users/dwmw2/openconnect.git \
+   file://0001-Makefile.am-add-missing-dependency.patch"
  
  S = "${WORKDIR}/git"
  



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

[oe] systemd support for recipes in meta-openembedded

[ChenQi]

Hi All,

In the next period, we are going to add systemd support (mostly systemd 
unit files) to recipes in meta-oe, meta-networking and meta-webserver.


If some of you are interested in doing such work, or you have being 
doing work in this area, please let us know so that we don't duplicate work.


Best Regards,
Chen Qi
--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH] xserver-nodm-init: remove recipe

[ChenQi]

On 12/18/2014 08:24 PM, Martin Jansa wrote:

On Thu, Dec 18, 2014 at 05:20:14PM +0800, qi.c...@windriver.com wrote:

From: Chen Qi 

Remove recipe as this recipe is in OE-core.

No, there is important diff which isn't in oe-core.


Hi Martin,

Is it possible to make the changes into OE-core?

I noticed this recipe in meta-oe because it caused build failures when I 
was building core-image-sato with debian package backend.



ERROR: Unable to install packages. Command 
'/buildarea2/chenqi/poky/build-systemd/tmp/sysroots/x86_64-linux/usr/bin/apt-get 
install --force-yes --allow-unauthenticated 
packagegroup-core-ssh-dropbear packagegroup-core-x11-sato-games 
packagegroup-core-x11-base packagegroup-core-boot psplash apt 
packagegroup-core-x11-sato packagegroup-base-extended dpkg' returned 100:

Reading package lists...
Building dependency tree...
Reading state information...
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
 packagegroup-core-x11-base : Depends: packagegroup-core-x11-utils but 
it is not going to be installed
W: Unable to read 
/buildarea2/chenqi/poky/build-systemd/tmp/work/qemux86-poky-linux/core-image-sato/1.0-r0/apt/preferences.d/ 
- DirectoryExists (2: No such file or directory)

E: Unable to correct problems, you have held broken packages.

ERROR: Function failed: do_rootfs
ERROR: Logfile of failure stored in: 
/buildarea2/chenqi/poky/build-systemd/tmp/work/qemux86-poky-linux/core-image-sato/1.0-r0/temp/log.do_rootfs.26735
ERROR: Task 7 
(/buildarea2/chenqi/poky/meta/recipes-sato/images/core-image-sato.bb, 
do_rootfs) failed with exit code '1'



The problem for this recipe is that it rdepends on 'xserver-common' 
which conflicts with 'x11-common'.

And we have:
meta/recipes-graphics/packagegroups/packagegroup-core-x11.bb:VIRTUAL-RUNTIME_xserver_common 
?= "x11-common"


If there's some blocking issues for this recipe to merge into OE-core, 
I'd like to patch it to make it rdepend on 
'VIRTUAL-RUNTIME_xserver_common' and set a default value in 
meta/conf/distro/include/default-providers.inc. What do you think?


P.S.
My build is on latest master of poky and meta-openembedded.

chenqi@pek-hostel-deb01:~/poky/meta-openembedded [0] $ git status
On branch master
Your branch is up-to-date with 'origin/master'.

nothing to commit, working directory clean
chenqi@pek-hostel-deb01:~/poky/meta-openembedded [0] $ cd ..
chenqi@pek-hostel-deb01:~/poky [0] $ git status
On branch master
Your branch is up-to-date with 'origin/master'.

nothing to commit, working directory clean

Best Regards,
Chen Qi


Signed-off-by: Chen Qi 
---
  .../xserver-nodm-init-2.0/gplv2-license.patch  | 353 -
  .../xserver-nodm-init-2.0/xserver-nodm |  51 ---
  .../xserver-nodm-init-2.0/xserver-nodm.conf|   1 -
  .../xserver-nodm-init-2.0/xserver-nodm.service |   9 -
  .../xserver-nodm-init/xserver-nodm-init_2.0.bb |  41 ---
  5 files changed, 455 deletions(-)
  delete mode 100644 
meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init-2.0/gplv2-license.patch
  delete mode 100755 
meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init-2.0/xserver-nodm
  delete mode 100644 
meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init-2.0/xserver-nodm.conf
  delete mode 100644 
meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init-2.0/xserver-nodm.service
  delete mode 100644 
meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init_2.0.bb

diff --git 
a/meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init-2.0/gplv2-license.patch
 
b/meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init-2.0/gplv2-license.patch
deleted file mode 100644
index ec93253..000
--- 
a/meta-oe/recipes-graphics/xserver-nodm-init/xserver-nodm-init-2.0/gplv2-license.patch
+++ /dev/null
@@ -1,353 +0,0 @@
-COPYING: add GPLv2 license file
-
-this is a local file recipe and the license file is missing.In order
-to pass the license checksum checking, the license file is needed. So
-this patch add the GPLv2 license file.
-
-Signed-off-by: Yu Ke 
-
-diff --git a/COPYING b/COPYING
-new file mode 100644
-index 000..d511905
 /dev/null
-+++ b/COPYING
-@@ -0,0 +1,339 @@
-+  GNU GENERAL PUBLIC LICENSE
-+ Version 2, June 1991
-+
-+ Copyright (C) 1989, 1991 Free Software Foundation, Inc.,
-+ 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
-+ Everyone is permitted to copy and distribute verbatim copies
-+ of this license document, but changing it is not allowed.
-+
-+  Preamble
-+
-+  The licenses for most software are de

Re: [oe] [meta-webserver][PATCH] webmin: fix failures in do_install task

[ChenQi]

On 12/03/2013 06:19 PM, Paul Eggleton wrote:

Hi Qi,

On Friday 22 November 2013 10:37:29 qi.c...@windriver.com wrote:

-export perl=perl
+export perl=${OECMAKE_PERLNATIVE_DIR}/perl

I'm afraid I don't like the look of this - OECMAKE_PERLNATIVE_DIR is set for
use by cmake, and perlnative.bbclass should be putting our perl binary into
PATH so it should be run anyway. Is this change actually necessary?

Cheers,
Paul



Please drop this patch. I found that the do_install task in webmin runs 
some perl scripts (e.g. postinstall.pl and enable-collections.pl) that 
are supposed to be run on target. I suspect webmin has never worked 
correctly on target as a result. I'm gonna propose another fix. The goal 
is to make it work on target.


Best Regards,
Chen Qi
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-webserver][PATCH] webmin: fix failures in do_install task

[ChenQi]

On 12/03/2013 07:02 PM, Paul Eggleton wrote:

On Tuesday 03 December 2013 18:45:37 ChenQi wrote:

On 12/03/2013 06:19 PM, Paul Eggleton wrote:

On Friday 22 November 2013 10:37:29 qi.c...@windriver.com wrote:

-export perl=perl
+export perl=${OECMAKE_PERLNATIVE_DIR}/perl

I'm afraid I don't like the look of this - OECMAKE_PERLNATIVE_DIR is set
for use by cmake, and perlnative.bbclass should be putting our perl
binary into PATH so it should be run anyway. Is this change actually
necessary?

Please drop this patch. I found that the do_install task in webmin runs
some perl scripts (e.g. postinstall.pl and enable-collections.pl) that
are supposed to be run on target. I suspect webmin has never worked
correctly on target as a result. I'm gonna propose another fix. The goal
is to make it work on target.

I can assure you that it has worked on the target; at least it did about 6
months ago (perhaps more by chance rather than design though).

Cheers,
Paul



Hi Paul,

Thanks for your information :)

I just tried it out on sato image with 'webmin' and 'midori' installed.
1) https://localhost:1
Cannot connect
2) http://localhost:1
Login page appeared, but I cannot login even if I provided the correct 
username and password.


I did this test on qemu, but I think the result should have nothing to 
do with qemu.


Best Regards,
Chen Qi
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe PATCH] busybox: Follow OE-Core update to 1.12.1 version

[ChenQi]

On 07/23/2013 08:54 PM, Paul Eggleton wrote:

On Tuesday 23 July 2013 09:46:36 Otavio Salvador wrote:

On Tue, Jul 23, 2013 at 9:41 AM, Paul Eggleton

 wrote:

On Tuesday 23 July 2013 09:09:37 Otavio Salvador wrote:

On Tue, Jul 23, 2013 at 6:44 AM, Paul Eggleton

 wrote:

On Monday 08 July 2013 13:35:21 Paul Eggleton wrote:

On Friday 05 July 2013 19:45:25 Eric Bénard wrote:

Le Fri, 05 Jul 2013 18:29:04 +0100,

Paul Eggleton  a écrit :

Could someone *please* figure out what this bbappend is needed for
and
get
it into OE-Core if it is still needed so we can drop it here?

it's only here to install /etc/default/busybox-syslog which contains
only one line : OPTIONS="-C64" which sets the size of the logbuffer.
So obviously this could go into OE-Core.

Yes but why is this needed? If we are to merge this into OE-Core we'll
need
to say why this change is needed in the commit message.

ping...

I am working in something else so I cannot work on busybox now. Feel
free to handle this.

The syslog support would be good to be kept.

The problem is it's not clear to me what we should do here. Surely having
this bbappend or not isn't the determining factor in whether we have
syslog support with systemd...?

With systemd we shouldn't provide it but for sysv we should provide it.

I still need to know what this is needed for. Surely it doesn't actually
enable syslog functionality, just adds one configuration value to it? If that's
the case, why is this configuration change needed?

Cheers,
Paul


Hi All,

FWIW, there are two bugs which might be related to this.
https://bugzilla.yoctoproject.org/show_bug.cgi?id=4837
https://bugzilla.yoctoproject.org/show_bug.cgi?id=4860

My opinion is to respect the same configuration file in both sysv and 
systemd systems.

The configuration file is /etc/syslog-startup.conf.

Best Regards,
Chen Qi
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-selinux][PATCH] README: update the Dependencies area

[ChenQi]

On 10/09/2013 08:21 PM, Joe MacDonald wrote:

[[oe] [meta-selinux][PATCH] README: update the Dependencies area] On 13.10.09 
(Wed 17:20) qi.c...@windriver.com wrote:


From: Chen Qi 

Add meta-oe to the dependencies as some recipes in meta-selinux may
require it. Here's a dependency chain:

['policycoreutils', 'libselinux', 'swig-native']

And swig is from meta-oe layer.

I'll take a look at this.  It shouldn't be necessary, meta-selinux
provides its own swig.

meta-selinux/recipes-devtools/swig
??? swig
?   ??? 0001-Use-proc-self-exe-for-swig-swiglib-on-non-Win32-plat.patch
??? swig_2.0.10.bb
??? swig.inc

Might've crept in during the update last week.

-J.


Thanks for pointing it out.
Yes. I checked again, meta-selinux has its own swig.
When I was verifying meta-selinux, I used wr-20130919 branch instead of 
the upstream-master branch.
On wr-20130919 branch, swig is not available. Is this intended? Do we 
need to update the README?


Best Regards,
Chen Qi


Signed-off-by: Chen Qi 
---
  README |7 +++
  1 file changed, 7 insertions(+)

diff --git a/README b/README
index 926d880..987f7ac 100644
--- a/README
+++ b/README
@@ -34,6 +34,13 @@ the Yocto Project website:
  
  http://www.yoctoproject.org
  
+For some recipes, the meta-oe layer is required:

+
+URI: git://github.com/openembedded/meta-oe.git
+subdirectory: meta-oe
+branch: master
+revision: HEAD
+
  
  Maintenance

  ---


___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel


___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH V2] gnuplot: upgrade to 5.0.5

[ChenQi]

On 11/04/2016 10:23 AM, Tim Orling wrote:

I would rather see a PACKAGECONFIG for qt5

On Tue, Nov 1, 2016 at 11:30 PM, Chen Qi  wrote:


Add '--disable-qt' to EXTRA_OECONF to fix do_compile failure.

Signed-off-by: Chen Qi 
---
  .../gnuplot-4.6.5/lua-loadlibs-configure-in-fix.patch| 16

  .../recipes-extended/gnuplot/gnuplot-4.6.5/subdirs.patch | 11 ---
  meta-oe/recipes-extended/gnuplot/gnuplot.inc |  2 +-
  .../gnuplot/gnuplot/lua-loadlibs-configure-in-fix.patch  | 16

  meta-oe/recipes-extended/gnuplot/gnuplot/subdirs.patch   | 11 +++
  meta-oe/recipes-extended/gnuplot/gnuplot_4.6.5.bb| 14
--
  meta-oe/recipes-extended/gnuplot/gnuplot_5.0.5.bb| 14
++
  7 files changed, 42 insertions(+), 42 deletions(-)
  delete mode 100644 meta-oe/recipes-extended/gnuplot/gnuplot-4.6.5/lua-
loadlibs-configure-in-fix.patch
  delete mode 100644 meta-oe/recipes-extended/
gnuplot/gnuplot-4.6.5/subdirs.patch
  create mode 100644 meta-oe/recipes-extended/gnuplot/gnuplot/lua-loadlibs-
configure-in-fix.patch
  create mode 100644 meta-oe/recipes-extended/gnuplot/gnuplot/subdirs.patch
  delete mode 100644 meta-oe/recipes-extended/gnuplot/gnuplot_4.6.5.bb
  create mode 100644 meta-oe/recipes-extended/gnuplot/gnuplot_5.0.5.bb

diff --git a/meta-oe/recipes-extended/gnuplot/gnuplot-4.6.5/lua-
loadlibs-configure-in-fix.patch b/meta-oe/recipes-extended/
gnuplot/gnuplot-4.6.5/lua-loadlibs-configure-in-fix.patch
deleted file mode 100644
index 23f2cd2..000
--- a/meta-oe/recipes-extended/gnuplot/gnuplot-4.6.5/lua-
loadlibs-configure-in-fix.patch
+++ /dev/null
@@ -1,16 +0,0 @@
-Index: gnuplot-4.6.5/configure.in
-===
 gnuplot-4.6.5.orig/configure.in
-+++ gnuplot-4.6.5/configure.in
-@@ -690,6 +690,11 @@ if test "${with_lua}" = yes ; then
-   fi
-
-   if test "$with_lua" != no; then
-+dnl check for dlopen/dl to fix loadlibs link failure
-+AC_CHECK_FUNC([dlopen], [],
-+  AC_CHECK_LIB([dl], [dlopen], DLOPEN_LIBS="-ldl"))
-+AC_SUBST(DLOPEN_LIBS)
-+LUA_LIBS="$LUA_LIBS $DLOPEN_LIBS"
- TERMLIBS="$TERMLIBS $LUA_LIBS"
- CPPFLAGS="$CPPFLAGS $LUA_CFLAGS"
-   else
diff --git a/meta-oe/recipes-extended/gnuplot/gnuplot-4.6.5/subdirs.patch
b/meta-oe/recipes-extended/gnuplot/gnuplot-4.6.5/subdirs.patch
deleted file mode 100644
index 8c7d1f0..000
--- a/meta-oe/recipes-extended/gnuplot/gnuplot-4.6.5/subdirs.patch
+++ /dev/null
@@ -1,11 +0,0 @@
 /tmp/Makefile.am   2007-08-23 13:10:15.560659023 +0200
-+++ gnuplot-4.2.0/Makefile.am  2007-08-23 13:10:34.961764629 +0200
-@@ -1,7 +1,7 @@
- ## Process this file with automake to produce Makefile.in -*-Makefile-*-
- AUTOMAKE_OPTIONS = foreign 1.2h
-
--SUBDIRS = config m4 term src docs $(LISPDIR) man demo tutorial share
-+SUBDIRS = config m4 term src  $(LISPDIR) man share
-
- EXTRA_DIST = BUGS CodeStyle Copyright FAQ GNUmakefile INSTALL
INSTALL.gnu \
- Makefile.maint PATCHLEVEL PGPKEYS PORTING README README.1ST README.exp \
diff --git a/meta-oe/recipes-extended/gnuplot/gnuplot.inc
b/meta-oe/recipes-extended/gnuplot/gnuplot.inc
index 69bc74b..d8d9a5f 100644
--- a/meta-oe/recipes-extended/gnuplot/gnuplot.inc
+++ b/meta-oe/recipes-extended/gnuplot/gnuplot.inc
@@ -17,7 +17,7 @@ PACKAGECONFIG[cairo] = "--with-cairo,--without-cairo,cairo
pango"
  PACKAGECONFIG[lua] = "--with-lua,--without-lua,lua"

  EXTRA_OECONF = "--with-readline=${STAGING_LIBDIR}/.. \
---without-lisp-files \
+--without-qt \


This would be more universal in a PACKAGECONFIG
e.g. PACKAGECONFIG[qt4] = "--with-qt,--without-qt,qt4-x11-free"



The problem is that it fails to compile with qt4. So I just disabled it.

Regards,
Chen Qi


  --without-tutorial \
  --disable-wxwidgets"

diff --git 
a/meta-oe/recipes-extended/gnuplot/gnuplot/lua-loadlibs-configure-in-fix.patch
b/meta-oe/recipes-extended/gnuplot/gnuplot/lua-loadlibs-
configure-in-fix.patch
new file mode 100644
index 000..23f2cd2
--- /dev/null
+++ b/meta-oe/recipes-extended/gnuplot/gnuplot/lua-loadlibs-
configure-in-fix.patch
@@ -0,0 +1,16 @@
+Index: gnuplot-4.6.5/configure.in
+===
+--- gnuplot-4.6.5.orig/configure.in
 gnuplot-4.6.5/configure.in
+@@ -690,6 +690,11 @@ if test "${with_lua}" = yes ; then
+   fi
+
+   if test "$with_lua" != no; then
++dnl check for dlopen/dl to fix loadlibs link failure
++AC_CHECK_FUNC([dlopen], [],
++  AC_CHECK_LIB([dl], [dlopen], DLOPEN_LIBS="-ldl"))
++AC_SUBST(DLOPEN_LIBS)
++LUA_LIBS="$LUA_LIBS $DLOPEN_LIBS"
+ TERMLIBS="$TERMLIBS $LUA_LIBS"
+ CPPFLAGS="$CPPFLAGS $LUA_CFLAGS"
+   else
diff --git a/meta-oe/recipes-extended/gnuplot/gnuplot/subdirs.patch
b/meta-oe/recipes-extended/gnuplot/gnuplot/subdirs.patch
new file mode 100644
index 000..8c7d1f0
--- /dev/null
+++ b/meta-oe/recipes-extended/gnupl

Re: [oe] [meta-oe][PATCH] rsyslog: use atomic builtins to avoid race issue

[ChenQi]

On 12/29/2016 07:20 AM, Khem Raj wrote:

On Tue, Dec 27, 2016 at 11:46 PM, Chen Qi  wrote:

rsyslog uses a small program to check if compiler has atomic builtins
or not. This does not work when cross compiling, resulting in potential
race issue. In fact, when running ptest on intel-x86-64 board, we have
met the race problem.

As our compiler gcc, has atomic builtins, we force ap_cv_atomic_builtins
to be 'yes' so that rsyslog could make use of the atomic builtins and avoid
the race issue.

Signed-off-by: Chen Qi 
---
  meta-oe/recipes-extended/rsyslog/rsyslog_8.22.0.bb | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/meta-oe/recipes-extended/rsyslog/rsyslog_8.22.0.bb 
b/meta-oe/recipes-extended/rsyslog/rsyslog_8.22.0.bb
index 4e2c122..27162e1 100644
--- a/meta-oe/recipes-extended/rsyslog/rsyslog_8.22.0.bb
+++ b/meta-oe/recipes-extended/rsyslog/rsyslog_8.22.0.bb
@@ -31,7 +31,7 @@ SRC_URI[sha256sum] = 
"06e2884181333dccecceaca82827ae24ca7a258b4fbf7b1e07a80d4caa

  inherit autotools pkgconfig systemd update-rc.d update-alternatives ptest

-EXTRA_OECONF += "--disable-generate-man-pages"
+EXTRA_OECONF += "--disable-generate-man-pages ap_cv_atomic_builtins=yes"

Does this also cover 64bit builtins ? secondly, a note for
architectures not supporting atomics to disable this would be good.


No. When looking at the source, I think it's not critical.
When I made this patch, my criteria is:
1) ptest of rsyslog does not hang because of race issue on intel-x86-64 
board
2) rsyslog works well on qemux86, qemux86-64, qemuarm, qemumips, 
qemumips64, qemuppc


I searched the doc and the internet, I could not find related 
information on which arch gcc does not support atomic builtins.


Best Regards,
Chen Qi


  # first line is default yes in configure
  PACKAGECONFIG ??= " \
--
2.8.3

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH 4/4] lvm2/libdevmapper: fix invoking lvcreate failed

[ChenQi]

I think I've got to reply. But frankly, I don't have any idea how to 
solve this circular dependency problem.


Normally we have two ways to solve circular dependency problems:
1) Add some new recipe to replace one in the chain
2) Break some recipe in the chain into multiple recipes, giving them 
different dependencies


In this systemd/cryptsetup/libdevmapper case, I don't think the above 
approaches would work.


Hope someone in the mailing list could give me some suggestions.

Best Regards,
Chen Qi

On 05/11/2018 10:48 PM, Chris Patterson wrote:

Sorry to resurrect such an old thread... :)

The dependency on udev brings back the c the libdevmapper circular
dependency when systemd is used with cryptsetup.  (systemd ->
cryptsetup -> libdevmapper -> udev|systemd).

Any thoughts on how to properly fix that (other than removing udev
from PACKAGECONFIG)?

On Fri, May 26, 2017 at 3:37 AM, Hongxu Jia  wrote:

Previously, libdevmapper is separated from lvm2 in the following commit:
---
commit 7f17da154372b5ec33947766783444961e8aa67c
Author: Chen Qi 
Date:   Mon Apr 24 16:06:41 2017 +0800

 lvm2: separate libdevmapper to avoid circular dependency
---

It caused invoking lvcreate failed:
---
root@qemux86-64:~# lvm pvcreate /dev/sda
   Physical volume "/dev/sda" successfully created.
root@qemux86-64:~# lvm vgcreate -s 4096K wrl /dev/sda
   Volume group "wrl" successfully created
root@qemux86-64:~# lvm lvcreate -n swap -L 2293648K -y --type linear wrl
   Rounding up size to full physical extent 2.19 GiB
   /dev/wrl/swap: not found: device not cleared
   Aborting. Failed to wipe start of new LV.
---

The lvm2 and libdevmapper should use the same PACKAGECONFIG
and EXTRA_OECONF.

And do not manually remove useless files.

Signed-off-by: Hongxu Jia 
---
  .../recipes-support/lvm2/libdevmapper_2.02.166.bb  | 21 ++---
  meta-oe/recipes-support/lvm2/lvm2.inc  | 26 +++-
  .../0001-explicitly-do-not-install-libdm.patch | 36 ++
  meta-oe/recipes-support/lvm2/lvm2_2.02.171.bb  | 31 ++-
  4 files changed, 65 insertions(+), 49 deletions(-)
  create mode 100644 
meta-oe/recipes-support/lvm2/lvm2/0001-explicitly-do-not-install-libdm.patch

diff --git a/meta-oe/recipes-support/lvm2/libdevmapper_2.02.166.bb 
b/meta-oe/recipes-support/lvm2/libdevmapper_2.02.166.bb
index fd1fc2c..3cf939d 100644
--- a/meta-oe/recipes-support/lvm2/libdevmapper_2.02.166.bb
+++ b/meta-oe/recipes-support/lvm2/libdevmapper_2.02.166.bb
@@ -5,27 +5,10 @@ SRC_URI[sha256sum] = 
"e120b066b85b224552efda40204488c5123de068725676fd6e5c8bc655

  DEPENDS += "autoconf-archive-native"

-PACKAGECONFIG = ""
-
-# Unset user/group to unbreak install.
-EXTRA_OECONF = "--with-user= \
---with-group= \
---enable-pkgconfig \
---with-usrlibdir=${libdir} \
-"
-
  TARGET_CC_ARCH += "${LDFLAGS}"

-do_install_append() {
-# Remove things unrelated to libdevmapper
-rm -rf ${D}${sysconfdir}
-for i in `ls ${D}${sbindir}/*`; do
-   if [ $i != ${D}${sbindir}/dmsetup ]; then
-   rm $i
-   fi
-done
-# Remove docs
-rm -rf ${D}${datadir}
+do_install() {
+oe_runmake 'DESTDIR=${D}' -C libdm install
  }

  RRECOMMENDS_${PN}_append_class-target = " lvm2-udevrules"
diff --git a/meta-oe/recipes-support/lvm2/lvm2.inc 
b/meta-oe/recipes-support/lvm2/lvm2.inc
index cb5a9e0..c2dac95 100644
--- a/meta-oe/recipes-support/lvm2/lvm2.inc
+++ b/meta-oe/recipes-support/lvm2/lvm2.inc
@@ -18,11 +18,18 @@ S = "${WORKDIR}/LVM2.${PV}"

  inherit autotools-brokensep pkgconfig systemd

+LVM2_PACKAGECONFIG = "dmeventd lvmetad"
+LVM2_PACKAGECONFIG_append_class-target = " \
+${@bb.utils.filter('DISTRO_FEATURES', 'selinux', d)} \
+thin-provisioning-tools \
+udev \
+"
+
  # odirect is always enabled because there currently is a bug in
  # lib/device/dev-io.c which prevents compiling without it. It is
  # better to stick to configurations that were actually tested by
  # upstream...
-PACKAGECONFIG_append = " odirect"
+PACKAGECONFIG ??= "odirect ${LVM2_PACKAGECONFIG}"

  PACKAGECONFIG[dmeventd] = "--enable-dmeventd,--disable-dmeventd"
  PACKAGECONFIG[lvmetad] = "--enable-lvmetad,--disable-lvmetad"
@@ -31,3 +38,20 @@ PACKAGECONFIG[readline] = 
"--enable-readline,--disable-readline,readline"
  PACKAGECONFIG[selinux] = "--enable-selinux,--disable-selinux,libselinux"
  PACKAGECONFIG[thin-provisioning-tools] = 
"--with-thin=internal,--with-thin=none,,thin-provisioning-tools"
  PACKAGECONFIG[udev] = "--enable-udev_sync --enable-udev_rules 
--with-udevdir=${nonarch_base_libdir}/udev/rules.d,--disable-udev_sync 
--disable-udev_rules,udev"
+
+# Unset user/group to unbreak install.
+EXTRA_OECONF = "--with-user= \
+--with-group= \
+--enable-realtime \
+--enable-applib \
+--enable-cmdlib \
+--enable-pkgconfig \
+--with-usrlibdir=${libdir} \
+--w

Re: [oe] [meta-oe][PATCH 1/2] rsyslog: fix configuration file for logrotate

[ChenQi]

On 08/04/2018 12:33 AM, Khem Raj wrote:

is this patch any different from last posted patch ?

https://patchwork.openembedded.org/patch/153125/


There's is no difference.
I sent it out again because the second patch also modifies the 
do_install function and I thought if I send them out together, it will 
be more convenient.


Best Regards,
Chen Qi


If not then there is no need to resend, if yes then please describe
the change and mark
patch as v2, v3 and so on. Save some time for reviewers
On Fri, Aug 3, 2018 at 3:02 AM  wrote:

From: Chen Qi 

The logrotate.rsyslog file uses 'cat /var/run/rsyslogd.pid' to
determine the pid of rsyslogd daemon. This is incorrect, because
rsyslogd may start in a way that does not create rsyslog.pid file.

Failing to send HUP signal to rsyslogd daemon will result in
unexpected behaviour. e.g. rsyslogd is writing to /var/log/messages-
instead of /var/log/messages.

We should really use pkill to send HUP signal to rsyslogd to ensure
correct behaviour.

Signed-off-by: Chen Qi 
---
  meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate | 4 ++--
  meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb | 1 +
  2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate 
b/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate
index 94ec517..5f8568f 100644
--- a/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate
+++ b/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate
@@ -9,7 +9,7 @@
  delaycompress
  compress
  postrotate
-   /bin/kill -HUP `cat /var/run/rsyslogd.pid 2> /dev/null` 2> 
/dev/null || true
+   @BINDIR@/pkill -HUP rsyslogd 2> /dev/null || true
  endscript
  }

@@ -34,6 +34,6 @@
  delaycompress
  sharedscripts
  postrotate
-   /bin/kill -HUP `cat /var/run/rsyslogd.pid 2> /dev/null` 2> 
/dev/null || true
+   @BINDIR@/pkill -HUP rsyslogd 2> /dev/null || true
  endscript
  }
diff --git a/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb 
b/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb
index 9737892..d90695a 100644
--- a/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb
+++ b/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb
@@ -125,6 +125,7 @@ do_install_append() {
  install -m 755 ${WORKDIR}/initscript ${D}${sysconfdir}/init.d/syslog
  install -m 644 ${WORKDIR}/rsyslog.conf ${D}${sysconfdir}/rsyslog.conf
  install -m 644 ${WORKDIR}/rsyslog.logrotate 
${D}${sysconfdir}/logrotate.d/logrotate.rsyslog
+sed -i -e "s#@BINDIR@#${bindir}#g" 
${D}${sysconfdir}/logrotate.d/logrotate.rsyslog
  }

  FILES_${PN} += "${bindir}"
--
1.9.1

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH 2/2] rsyslog: add PACKAGECONFIG for imjournal

[ChenQi]

On 08/04/2018 12:27 AM, Khem Raj wrote:

On Fri, Aug 3, 2018 at 3:03 AM  wrote:

From: Chen Qi 

Add PACKAGECONFIG for imjournal, and add corresponding configuration
to /etc/rsyslog.conf when it's enabled. Default to disable it.

Signed-off-by: Chen Qi 
---
  meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb | 5 +
  1 file changed, 5 insertions(+)

diff --git a/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb 
b/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb
index d90695a..1bcbde1 100644
--- a/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb
+++ b/meta-oe/recipes-extended/rsyslog/rsyslog_8.36.0.bb
@@ -67,6 +67,7 @@ PACKAGECONFIG[imfile] = "--enable-imfile,--disable-imfile,,"
  PACKAGECONFIG[snmp] = "--enable-snmp,--disable-snmp,net-snmp,"
  PACKAGECONFIG[gnutls] = "--enable-gnutls,--disable-gnutls,gnutls,"
  PACKAGECONFIG[systemd] = 
"--with-systemdsystemunitdir=${systemd_unitdir}/system/,--without-systemdsystemunitdir,systemd,"
+PACKAGECONFIG[imjournal] = "--enable-imjournal,--disable-imjournal,"
  PACKAGECONFIG[mysql] = "--enable-mysql,--disable-mysql,mysql5,"
  PACKAGECONFIG[postgresql] = "--enable-pgsql,--disable-pgsql,postgresql,"
  PACKAGECONFIG[libdbi] = "--enable-libdbi,--disable-libdbi,libdbi,"
@@ -126,6 +127,10 @@ do_install_append() {
  install -m 644 ${WORKDIR}/rsyslog.conf ${D}${sysconfdir}/rsyslog.conf
  install -m 644 ${WORKDIR}/rsyslog.logrotate 
${D}${sysconfdir}/logrotate.d/logrotate.rsyslog
  sed -i -e "s#@BINDIR@#${bindir}#g" 
${D}${sysconfdir}/logrotate.d/logrotate.rsyslog
+
+if ${@bb.utils.contains('PACKAGECONFIG', 'imjournal', 'true', 'false', 
d)}; then
+echo '$ModLoad imjournal' >> ${D}${sysconfdir}/rsyslog.conf
+fi

perhaps its time to add something like

$IncludeConfig /etc/rsyslog.d/*.conf

in main etc/rsyslog.conf and then such fragments can be places as
individual config
files inside that directory


I agree. I'll send out V2.

Best Regards,
Chen Qi


  }

  FILES_${PN} += "${bindir}"
--
1.9.1

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-networking][PATCH] postfix: fix pkg_postinst_${PN}

[ChenQi]

On 09/13/2018 04:23 PM, mingli...@windriver.com wrote:

From: Mingli Yu 

"exit 0" will break the postinst logic
below the line "exit 0" such as:
===
update-alternatives --install /usr/sbin/sendmail sendmail 
/usr/sbin/sendmail.postfix 120
update-alternatives --install /usr/bin/mailq mailq /usr/bin/mailq.postfix 120
update-alternatives --install /usr/bin/newaliases newaliases 
/usr/bin/newaliases.postfix 120
===

It's enough to call "$INTERCEPT_DIR/postinst_intercept delay_to_first_boot ${PKG} 
mlprefix=${MLPREFIX}"
, so remove it to guarantee the postinstall logic is correct

Signed-off-by: Mingli Yu 
---
  meta-networking/recipes-daemons/postfix/postfix.inc | 1 -
  1 file changed, 1 deletion(-)

diff --git a/meta-networking/recipes-daemons/postfix/postfix.inc 
b/meta-networking/recipes-daemons/postfix/postfix.inc
index 49f9f7b1f..71d7ef5df 100644
--- a/meta-networking/recipes-daemons/postfix/postfix.inc
+++ b/meta-networking/recipes-daemons/postfix/postfix.inc
@@ -246,7 +246,6 @@ pkg_postinst_${PN} () {
  # This can fail depending on host setup
  if ! newaliases -C $D/etc/postfix/main.cf -oA$D/etc/aliases; then
  $INTERCEPT_DIR/postinst_intercept delay_to_first_boot ${PKG} 
mlprefix=${MLPREFIX}
-exit 0
  fi
  touch $D/etc/postfix/virtual_alias
  postmap -c $D/etc/postfix $D/etc/postfix/virtual_alias

I guess you need to use the if...else... block here.

Regards,
Chen Qi
--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH 1/2] lvm2: separate libdevmapper to avoid circular dependency

[ChenQi]

On 04/20/2017 03:08 PM, Martin Jansa wrote:

On Tue, Apr 18, 2017 at 05:52:46PM +0800, qi.c...@windriver.com wrote:

From: Chen Qi 

If we enable 'cryptsetup' PACKAGECONFIG for systemd, there would be
circular dependency issue among 'systemd', 'cryptsetup' and 'lvm2'.

cryptsetup only needs libdevmapper from lvm2. So we separte libdevmapper
into a new recipe to solve this circular dependency issue.

Seems to break multipath-tools build:
http://errors.yoctoproject.org/Errors/Details/140512/

Also in one of my builds it failed to install with:
| rm: cannot remove lvm2/2.02.166-r0/image/usr/lib/pkgconfig/devmapper.pc: No 
such file or directory

because my lvm2 build didn't have "--enable-pkgconfig" option (we're
using only small part of lvm2), maybe it would be better to use rm -f
when removing devmapper files.


Thanks for pointing out these problem.
I'll fix them and send out a new patch series.

Best Regards,
Chen Qi


Signed-off-by: Chen Qi 
---
  .../0001-Avoid-bashisms-in-init-scripts.patch  | 182 
  ...001-implement-libc-specific-reopen_stream.patch |  79 +
  ...0002-Guard-use-of-mallinfo-with-__GLIBC__.patch |  30 ++
  ...cntl.h-for-O_-defines-and-fcntl-signature.patch |  30 ++
  ...0004-tweak-MODPROBE_CMD-for-cross-compile.patch |  38 +++
  meta-oe/recipes-support/lvm2/files/lvm.conf| 326 +
  .../recipes-support/lvm2/libdevmapper_2.02.166.bb  |  31 ++
  meta-oe/recipes-support/lvm2/lvm2.inc  |  75 -
  .../lvm2/0001-Avoid-bashisms-in-init-scripts.patch | 182 
  ...001-implement-libc-specific-reopen_stream.patch |  79 -
  ...0002-Guard-use-of-mallinfo-with-__GLIBC__.patch |  30 --
  ...cntl.h-for-O_-defines-and-fcntl-signature.patch |  30 --
  ...0004-tweak-MODPROBE_CMD-for-cross-compile.patch |  38 ---
  meta-oe/recipes-support/lvm2/lvm2/lvm.conf | 326 -
  meta-oe/recipes-support/lvm2/lvm2_2.02.166.bb  |  73 +
  15 files changed, 789 insertions(+), 760 deletions(-)
  create mode 100644 
meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch
  create mode 100644 
meta-oe/recipes-support/lvm2/files/0001-implement-libc-specific-reopen_stream.patch
  create mode 100644 
meta-oe/recipes-support/lvm2/files/0002-Guard-use-of-mallinfo-with-__GLIBC__.patch
  create mode 100644 
meta-oe/recipes-support/lvm2/files/0003-include-fcntl.h-for-O_-defines-and-fcntl-signature.patch
  create mode 100644 
meta-oe/recipes-support/lvm2/files/0004-tweak-MODPROBE_CMD-for-cross-compile.patch
  create mode 100644 meta-oe/recipes-support/lvm2/files/lvm.conf
  create mode 100644 meta-oe/recipes-support/lvm2/libdevmapper_2.02.166.bb
  delete mode 100644 
meta-oe/recipes-support/lvm2/lvm2/0001-Avoid-bashisms-in-init-scripts.patch
  delete mode 100644 
meta-oe/recipes-support/lvm2/lvm2/0001-implement-libc-specific-reopen_stream.patch
  delete mode 100644 
meta-oe/recipes-support/lvm2/lvm2/0002-Guard-use-of-mallinfo-with-__GLIBC__.patch
  delete mode 100644 
meta-oe/recipes-support/lvm2/lvm2/0003-include-fcntl.h-for-O_-defines-and-fcntl-signature.patch
  delete mode 100644 
meta-oe/recipes-support/lvm2/lvm2/0004-tweak-MODPROBE_CMD-for-cross-compile.patch
  delete mode 100644 meta-oe/recipes-support/lvm2/lvm2/lvm.conf

diff --git 
a/meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch 
b/meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch
new file mode 100644
index 000..e86ab25
--- /dev/null
+++ 
b/meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch
@@ -0,0 +1,182 @@
+From 916ea0c70fd063ab7b81f16fd917a75dc02edf4f Mon Sep 17 00:00:00 2001
+From: Peter Kjellerstedt 
+Date: Fri, 17 Mar 2017 03:18:28 +0100
+Subject: [PATCH] Avoid bashisms in init scripts
+
+Upstream-Status: Inappropriate
+
+Signed-off-by: Peter Kjellerstedt 
+---
+ scripts/blk_availability_init_red_hat.in  | 4 ++--
+ scripts/clvmd_init_red_hat.in | 6 +++---
+ scripts/cmirrord_init_red_hat.in  | 4 ++--
+ scripts/lvm2_cluster_activation_red_hat.sh.in | 4 ++--
+ scripts/lvm2_lvmetad_init_red_hat.in  | 4 ++--
+ scripts/lvm2_lvmpolld_init_red_hat.in | 4 ++--
+ scripts/lvm2_monitoring_init_red_hat.in   | 4 ++--
+ scripts/lvm2_monitoring_init_rhel4| 4 ++--
+ 8 files changed, 17 insertions(+), 17 deletions(-)
+
+diff --git a/scripts/blk_availability_init_red_hat.in 
b/scripts/blk_availability_init_red_hat.in
+index a84ffe7..6b855b7 100644
+--- a/scripts/blk_availability_init_red_hat.in
 b/scripts/blk_availability_init_red_hat.in
+@@ -1,4 +1,4 @@
+-#!/bin/bash
++#!/bin/sh
+ #
+ # Copyright (C) 2012 Red Hat, Inc. All rights reserved.
+ #
+@@ -53,6 +53,6 @@ case "$1" in
+   status)
+   ;;
+   *)
+-  echo $"Usage: $0 {start|stop|status}"
++  echo "Usage: $0 {start|stop|status}"
+   ;;
+ esac
+diff --git a/scripts/clvmd_init_red_hat.in b/scripts/clvmd_init_red_hat.in
+index d7f3392..abc8

Re: [oe] [meta-oe][PATCH 1/3] lvm2: separate libdevmapper to avoid circular dependency

[ChenQi]

On 04/24/2017 10:21 PM, Martin Jansa wrote:

I'm still seeing some issues with this applied:

ERROR: lvm2-2.02.166-r0 do_package: QA Issue: lvm2: Files/directories were 
installed but not shipped in any package:
   /usr/lib
Please set FILES such that these items are packaged. Alternatively if they are 
unneeded, avoid installing them or delete them within do_install.
ERROR: libdevmapper-2.02.166-r0 do_packagedata: The recipe libdevmapper is 
trying to install files into a shared area when those files already exist. 
Those files and their manifest location are:
   pkgdata/mach/runtime/libdevmapper
 (matched in manifest-mach-lvm2.packagedata)
   pkgdata/mach/runtime/libdevmapper-dev
 (matched in manifest-mach-lvm2.packagedata)
   pkgdata/mach/runtime/libdevmapper.packaged
 (matched in manifest-mach-lvm2.packagedata)
   pkgdata/mach/shlibs2/libdevmapper.list
 (matched in manifest-mach-lvm2.packagedata)
Please verify which recipe should provide the above files.
This was in incremental build, so the second issue might not be reproducible in 
clean build, but still something to confirm before merge to master.




Hi Martin,

I did see the second issue, and I can confirm it's related to 
incremental build. I think we can safely ignore this issue.


However, for the first issue, I haven't seen one. I tried to reproduce 
it today, but no luck.

Is there some way to reproduce it?

Best Regards,
Chen Qi

On Mon, Apr 24, 2017 at 10:06 AM, > wrote:


From: Chen Qi mailto:qi.c...@windriver.com>>

If we enable 'cryptsetup' PACKAGECONFIG for systemd, there would be
circular dependency issue among 'systemd', 'cryptsetup' and 'lvm2'.

cryptsetup only needs libdevmapper from lvm2. So we separte
libdevmapper
into a new recipe to solve this circular dependency issue.

Signed-off-by: Chen Qi mailto:qi.c...@windriver.com>>
---
 .../0001-Avoid-bashisms-in-init-scripts.patch  | 182 
 ...001-implement-libc-specific-reopen_stream.patch | 79 +
 ...0002-Guard-use-of-mallinfo-with-__GLIBC__.patch | 30 ++
 ...cntl.h-for-O_-defines-and-fcntl-signature.patch | 30 ++
 ...0004-tweak-MODPROBE_CMD-for-cross-compile.patch | 38 +++
 meta-oe/recipes-support/lvm2/files/lvm.conf| 326
+
 .../recipes-support/lvm2/libdevmapper_2.02.166.bb
  |  31 ++
 meta-oe/recipes-support/lvm2/lvm2.inc  | 75 -
 .../lvm2/0001-Avoid-bashisms-in-init-scripts.patch | 182 
 ...001-implement-libc-specific-reopen_stream.patch | 79 -
 ...0002-Guard-use-of-mallinfo-with-__GLIBC__.patch | 30 --
 ...cntl.h-for-O_-defines-and-fcntl-signature.patch | 30 --
 ...0004-tweak-MODPROBE_CMD-for-cross-compile.patch | 38 ---
 meta-oe/recipes-support/lvm2/lvm2/lvm.conf | 326
-
 meta-oe/recipes-support/lvm2/lvm2_2.02.166.bb
  |  73 +
 15 files changed, 789 insertions(+), 760 deletions(-)
 create mode 100644
meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch
 create mode 100644

meta-oe/recipes-support/lvm2/files/0001-implement-libc-specific-reopen_stream.patch
 create mode 100644

meta-oe/recipes-support/lvm2/files/0002-Guard-use-of-mallinfo-with-__GLIBC__.patch
 create mode 100644

meta-oe/recipes-support/lvm2/files/0003-include-fcntl.h-for-O_-defines-and-fcntl-signature.patch
 create mode 100644

meta-oe/recipes-support/lvm2/files/0004-tweak-MODPROBE_CMD-for-cross-compile.patch
 create mode 100644 meta-oe/recipes-support/lvm2/files/lvm.conf
 create mode 100644
meta-oe/recipes-support/lvm2/libdevmapper_2.02.166.bb

 delete mode 100644
meta-oe/recipes-support/lvm2/lvm2/0001-Avoid-bashisms-in-init-scripts.patch
 delete mode 100644

meta-oe/recipes-support/lvm2/lvm2/0001-implement-libc-specific-reopen_stream.patch
 delete mode 100644

meta-oe/recipes-support/lvm2/lvm2/0002-Guard-use-of-mallinfo-with-__GLIBC__.patch
 delete mode 100644

meta-oe/recipes-support/lvm2/lvm2/0003-include-fcntl.h-for-O_-defines-and-fcntl-signature.patch
 delete mode 100644

meta-oe/recipes-support/lvm2/lvm2/0004-tweak-MODPROBE_CMD-for-cross-compile.patch
 delete mode 100644 meta-oe/recipes-support/lvm2/lvm2/lvm.conf

diff --git

a/meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch

b/meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch
new file mode 100644
index 000..e86ab25
--- /dev/null
+++

b/meta-oe/recipes-support/lvm2/files/0001-Avoid-bashisms-in-init-scripts.patch
@@ -0,0 +1,182 @@
+From 916ea0c70fd063ab7b81f16fd917a75dc02edf4f Mon Sep 17 00:00:00
2001
+From: Peter Kjellerstedt mailto:p...@axis.com>>
+Date: Fri, 17 Mar 2

Re: [oe] [meta-filesystems][PATCH] xfsdump: upgrade to 3.1.6 and remove from blacklist

[ChenQi]

Sorry for my carelessness. I'll fix the problem ASAP and sent out a new 
patchset.


Best Regards,
Chen Qi

On 04/27/2017 10:44 PM, Martin Jansa wrote:

And it still fails:
http://errors.yoctoproject.org/Errors/Details/141368/

On Thu, Apr 27, 2017 at 8:37 AM, > wrote:


From: Robert Yang mailto:liezhi.y...@windriver.com>>

The 3.1.4 doesn't work with xfsprogs 4.7:
| inv_files.o: In function `inv_dirpath':
|xfsdump-3.1.4/invutil/inv_files.c:53: undefined reference to `ASSERT'
| inv_files.o: In function `inv_fstab':
|xfsdump-3.1.4/invutil/inv_files.c:60: undefined reference to `ASSERT'
| inv_files.o: In function `inv_lockfile':
|xfsdump-3.1.4/invutil/inv_files.c:67: undefined reference to `ASSERT'
| inv_files.o: In function `inv_basepath':
|xfsdump-3.1.4/invutil/inv_files.c:74: undefined reference to `ASSERT'
| collect2: error: ld returned 1 exit status
| make[2]: *** [../include/buildrules:33: xfsinvutil] Error 1

Upgrade to 3.1.6, and remove it from blacklist.

Signed-off-by: Robert Yang mailto:liezhi.y...@windriver.com>>
---
 .../recipes-utils/xfsdump/{xfsdump_3.1.4.bb
 => xfsdump_3.1.6.bb
} | 6 ++
 1 file changed, 2 insertions(+), 4 deletions(-)
 rename meta-filesystems/recipes-utils/xfsdump/{xfsdump_3.1.4.bb
 => xfsdump_3.1.6.bb
} (79%)

diff --git
a/meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.4.bb

b/meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.6.bb

similarity index 79%
rename from
meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.4.bb

rename to meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.6.bb

index d5ad0b0b2..fac9d47e8 100644
--- a/meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.4.bb

+++ b/meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.6.bb

@@ -13,8 +13,8 @@ DEPENDS = "xfsprogs attr"
 SRC_URI =
"ftp://oss.sgi.com/projects/xfs/cmd_tars/${BPN}-${PV}.tar.gz

\
 file://remove-install-as-user.patch \
 "
-SRC_URI[md5sum] = "a8b1761be5feb363131e7b506639ad4c"
-SRC_URI[sha256sum] =
"570eafd0721515bdd79cb0e295b701d49cdf81e71a0a0ff0df6d4c5cc1960943"
+SRC_URI[md5sum] = "50353cd4f4b435685955363e6044f4d1"
+SRC_URI[sha256sum] =
"7f78c11ca527477d90e5e62b0778f3ad96f2b71c19173044e9aca9515fff42d0"

 inherit autotools-brokensep

@@ -34,5 +34,3 @@ do_install () {
 oe_runmake install
 oe_runmake install-dev
 }
-
-PNBLACKLIST[xfsdump] ?= "Depends on broken xfsprogs - the recipe
will be removed on 2017-09-01 unless the issue is fixed"
--
2.11.0

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org

http://lists.openembedded.org/mailman/listinfo/openembedded-devel





--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-filesystems][PATCH 1/2] xfsdump: 3.1.4 -> 3.1.6

[ChenQi]

On 05/04/2017 05:55 PM, Martin Jansa wrote:

Have you kept the old version intentionally?



NO. That's a mistake when using 'git format-patch', I'll send out V2.
Thanks for reminding me about this.

Best Regards,
Chen Qi

On Thu, May 4, 2017 at 11:32 AM, Chen Qi > wrote:


From: Robert Yang mailto:liezhi.y...@windriver.com>>

The 3.1.4 doesn't work with xfsprogs 4.7:
| inv_files.o: In function `inv_dirpath':
|xfsdump-3.1.4/invutil/inv_files.c:53: undefined reference to `ASSERT'
| inv_files.o: In function `inv_fstab':
|xfsdump-3.1.4/invutil/inv_files.c:60: undefined reference to `ASSERT'
| inv_files.o: In function `inv_lockfile':
|xfsdump-3.1.4/invutil/inv_files.c:67: undefined reference to `ASSERT'
| inv_files.o: In function `inv_basepath':
|xfsdump-3.1.4/invutil/inv_files.c:74: undefined reference to `ASSERT'
| collect2: error: ld returned 1 exit status
| make[2]: *** [../include/buildrules:33: xfsinvutil] Error 1

Signed-off-by: Robert Yang mailto:liezhi.y...@windriver.com>>
---
 .../recipes-utils/xfsdump/xfsdump_3.1.6.bb
 | 38 ++
 1 file changed, 38 insertions(+)
 create mode 100644
meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.6.bb


diff --git
a/meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.6.bb

b/meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.6.bb

new file mode 100644
index 0..8f942ba7c
--- /dev/null
+++ b/meta-filesystems/recipes-utils/xfsdump/xfsdump_3.1.6.bb

@@ -0,0 +1,38 @@
+SUMMARY = "XFS Filesystem Dump Utility"
+DESCRIPTION = "The xfsdump package contains xfsdump, xfsrestore
and a \
+   number of other utilities for administering XFS
filesystems.\
+   xfsdump examines files in a filesystem, determines
which \
+   need to be backed up, and copies those files to a \
+   specified disk, tape or other storage medium."
+HOMEPAGE = "http://oss.sgi.com/projects/xfs
"
+SECTION = "base"
+LICENSE = "GPLv2"
+LIC_FILES_CHKSUM =
"file://doc/COPYING;md5=15c832894d10ddd00dfcf57bee490ecc"
+DEPENDS = "xfsprogs attr"
+
+SRC_URI =
"ftp://oss.sgi.com/projects/xfs/cmd_tars/${BPN}-${PV}.tar.gz

\
+file://remove-install-as-user.patch \
+"
+SRC_URI[md5sum] = "50353cd4f4b435685955363e6044f4d1"
+SRC_URI[sha256sum] =
"7f78c11ca527477d90e5e62b0778f3ad96f2b71c19173044e9aca9515fff42d0"
+
+inherit autotools-brokensep
+
+PARALLEL_MAKE = ""
+PACKAGECONFIG ??= ""
+PACKAGECONFIG[gettext] =
"--enable-gettext=yes,--enable-gettext=no,gettext"
+
+EXTRA_OEMAKE += "'LIBTOOL=${HOST_SYS}-libtool' V=1"
+
+do_configure () {
+export DEBUG="-DNDEBUG"
+oe_runconf
+}
+
+do_install () {
+export DIST_ROOT=${D}
+oe_runmake install
+oe_runmake install-dev
+}
+
+PNBLACKLIST[xfsdump] ?= "Depends on broken xfsprogs - the recipe
will be removed on 2017-09-01 unless the issue is fixed"
--
2.11.0

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org

http://lists.openembedded.org/mailman/listinfo/openembedded-devel





--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH] hwloc: add several PACKAGECONFIG items

[ChenQi]

On 06/23/2017 06:14 PM, Martin Jansa wrote:

Should numactl be enabled by default?

numactl.bb  currently marks it ARM incompatible:
# ARM does not currently support NUMA
COMPATIBLE_HOST = "^((?!arm).*)$"

and enabling the PACKAGECONFIG here by default would extend the same 
restriction to hwloc as well.





Thanks for pointing out the problem.
I'll send out V2.

Best Regards,
Chen Qi

On Thu, Jun 22, 2017 at 8:46 AM, Chen Qi > wrote:


Add PACKAGECONFIG items to allow for more flexiblity.

Signed-off-by: Chen Qi mailto:qi.c...@windriver.com>>
---
 meta-oe/recipes-extended/hwloc/hwloc_1.11.5.bb
 | 10 --
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/meta-oe/recipes-extended/hwloc/hwloc_1.11.5.bb

b/meta-oe/recipes-extended/hwloc/hwloc_1.11.5.bb

index e067f2799..16cd507e6 100644
--- a/meta-oe/recipes-extended/hwloc/hwloc_1.11.5.bb

+++ b/meta-oe/recipes-extended/hwloc/hwloc_1.11.5.bb

@@ -13,10 +13,16 @@ SRC_URI[sha256sum] =
"95d80286dfe658a3f79e2ac90698782bb36e5504f4bac1bba2394ba14d

 inherit autotools

-DEPENDS += "cairo ncurses udev libxml2 zlib libpciaccess"
-DEPENDS += "${@bb.utils.contains('DISTRO_FEATURES', 'x11',
'virtual/libx11', '', d)}"
+DEPENDS += "ncurses udev zlib"
 DEPENDS += "${@bb.utils.contains('DISTRO_FEATURES', 'selinux',
'libselinux', '', d)}"

+PACKAGECONFIG ?= "numactl pci libxml2
${@bb.utils.contains('DISTRO_FEATURES', 'x11', 'x11', '', d)}"
+
+PACKAGECONFIG[numactl] =
"--enable-libnuma,--disable-libnuma,numactl,numactl"
+PACKAGECONFIG[libxml2] =
"--enable-libxml2,--disable-libxml2,libxml2,libxml2"
+PACKAGECONFIG[x11] = "--with-x,--without-x,virtual/libx11
cairo,cairo"
+PACKAGECONFIG[pci] =
"--enable-pci,--disable-pci,libpciaccess,libpciaccess"
+
 # Split hwloc library into separate subpackage
 PACKAGES_prepend = " lib${PN} "
 FILES_lib${PN} += "${libdir}/lib${PN}.so*"
--
2.11.0

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org

http://lists.openembedded.org/mailman/listinfo/openembedded-devel





--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH] rsyslog: CVE-2015-3243

[ChenQi]

On 08/20/2017 10:51 AM, Zhixiong Chi wrote:

rsyslog uses weak permissions for generating log files, which allows
local users to obtain sensitive information by reading files in
/var/log/cron.log

We add "create 0600 root root" to the /etc/logrotate.d/syslog file,
this will ensure the file is created with permissions when logrotate
runs. It is also recommended that users manually set the permissions
on existing or newly installed log files in order to prevent access
by untrusted users.
https://bugzilla.redhat.com/show_bug.cgi?id=1232826

CVE: CVE-2015-3243

Signed-off-by: Zhixiong Chi 
---
  meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate | 3 +++
  1 file changed, 3 insertions(+)

diff --git a/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate 
b/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate
index 94ec517..7960815 100644
--- a/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate
+++ b/meta-oe/recipes-extended/rsyslog/rsyslog/rsyslog.logrotate
@@ -23,6 +23,9 @@
  /var/log/user.log
  /var/log/lpr.log
  /var/log/cron.log
+{
+create 0600 root root
+}
  /var/log/debug
  /var/log/messages
  {



Hi Zhixiong,

I also did some testing about this issue.

We use '0640' for these log files, owner is root and group is adm. So 
they are not world readable.


And I also tried logroate command on target to recreate these log files. 
They are created with 0640 file permission. (I checked the conf files, 
not sure why 0640 is used by default.) You could double check it if you 
like.


(I used 'logroate -f /etc/logroate.conf' command to do the test.)

P.S. Even if we want to do something, we should use 'create 0640 root adm'.

Best Regards,

Chen Qi

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-networking][PATCH] openconnect: fix parallel compilation failure

[ChenQi]

On 08/23/2017 06:46 PM, Jussi Kukkonen wrote:
On 23 August 2017 at 13:08, > wrote:


From: Chen Qi mailto:qi.c...@windriver.com>>

This patch solves the following error which appears now and then
because of
parallel building.

  ../git/main.c:78:21: fatal error: version.c: No such file or
directory

version.c is generated at build time. And it's needed by some
objects. So
it needs to be generated first to avoid compilation failure.

This patch fix the problem by generating version.c first before
building others.

Signed-off-by: Chen Qi mailto:qi.c...@windriver.com>>
---
 meta-networking/recipes-connectivity/openconnect/openconnect_git.bb
 | 4 
 1 file changed, 4 insertions(+)

diff --git
a/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb

b/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb

index 3a1f0f3..4242aa5 100644
---
a/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb

+++
b/meta-networking/recipes-connectivity/openconnect/openconnect_git.bb

@@ -25,3 +25,7 @@ inherit autotools pkgconfig

 EXTRA_OECONF +=
"--with-vpnc-script=${SYSROOT_DESTDIR}${sysconfdir}/vpnc/vpnc-script \
  --disable-static"
+
+do_compile_prepend() {
+make version.c
+}


Upstream has had a proper Makefile.am fix for this since August 2016. 
I think the recipe should just be upgraded -- VPN software from 2015 
is a bad idea.


Jussi



Thanks a lot. I'll send out a patch to upgrade it.

Best Regards,

Chen Qi

--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH] gnome-common: add PACKAGECONFIG for autoconf-archive

[ChenQi]

ping

On 10/24/2017 04:19 PM, qi.c...@windriver.com wrote:

From: Chen Qi 

Add PACKAGECONFIG for autoconf-archive. Default to enable it to avoid
m4 file conflicts between gnome-common and autoconf-archive.

e.g.
IMAGE_INSTALL_append = " gnome-common autoconf-archive-dev"
bitbake 

We enable autoconf-archive PACKAGECONFIG to avoid the following error.

Error: Transaction check error:
   file /usr/share/aclocal/ax_check_enable_debug.m4 conflicts between attempted 
installs of gnome-common-3.18.0-r0.noarch and 
autoconf-archive-dev-2016.09.16-r0.core2_64
   file /usr/share/aclocal/ax_code_coverage.m4 conflicts between attempted 
installs of gnome-common-3.18.0-r0.noarch and 
autoconf-archive-dev-2016.09.16-r0.core2_64

Signed-off-by: Chen Qi 
---
  meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb | 4 
  1 file changed, 4 insertions(+)

diff --git a/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb 
b/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
index 06f3bb3..d874164 100644
--- a/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
+++ b/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
@@ -14,6 +14,10 @@ SRC_URI[archive.sha256sum] = 
"22569e370ae755e04527b76328befc4c73b62bfd4a572499fd
  EXTRA_AUTORECONF = ""
  DEPENDS = ""
  
+# Default to enable autoconf-archive to avoid conflicts

+PACKAGECONFIG ??= "autoconf-archive"
+PACKAGECONFIG[autoconf-archive] = "--with-autoconf-archive, 
--without-autoconf-archive, autoconf-archive"
+
  FILES_${PN} += "${datadir}/aclocal"
  FILES_${PN}-dev = ""
  



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH] gnome-common: add PACKAGECONFIG for autoconf-archive

[ChenQi]

Hi Martin,

I sent out patch to oe-core to fix the problem you mentioned.
Now oe-core has merged the following patch. Could you please help 
merging this patch?


commit a9be3c445a047bc6fadf65b0d0f874c86c593e5d
Author: Chen Qi 
Date:   Mon Nov 13 13:26:22 2017 +0800

autoconf-archive: inherit allarch and fix package splitting

autoconf-archive is a set of common m4 macros, it should be allarch,
just like it is on other distros.

Also fix package splitting. With current package splitting,
autoconf-archive itself contains almost nothing, the m4 macros are
contained in its dev package. However, autoconf-archive should
contain those m4 marcos itself.

(From OE-Core rev: 8cbf5523f002e8ae0d8123755d2cbc3e8b0138ac)

Signed-off-by: Chen Qi 
Signed-off-by: Ross Burton 
Signed-off-by: Richard Purdie 

Best Regards,
Chen Qi

On 11/09/2017 05:16 PM, Martin Jansa wrote:
gnome-common is marked as allarch, autoconf-archive isn't. So 
gnome-common cannot depend on it as it is now.


This change causes:

 === Comparing signatures for task do_configure.sigdata between 
qemux86 and qemux86copy ===
ERROR: gnome-common different signature for task do_configure.sigdata 
between qemux86 and qemux86copy

runtaskdeps changed:
gnome-common/gnome-common_3.18.0.bb.do_prepare_recipe_sysroot with 
hash f3130e0be68211325008e37b3d49460b

 changed to
gnome-common/gnome-common_3.18.0.bb.do_prepare_recipe_sysroot with 
hash 0360cfa01ddf508a7670af2fcb306e20
Hash for dependent task 
gnome-common/gnome-common_3.18.0.bb.do_prepare_recipe_sysroot changed 
from f3130e0be68211325008e37b3d49460b to 0360cfa01ddf508a7670af2fcb306e20



On Tue, Nov 7, 2017 at 3:10 AM, ChenQi <mailto:qi.c...@windriver.com>> wrote:


ping


On 10/24/2017 04:19 PM, qi.c...@windriver.com
<mailto:qi.c...@windriver.com> wrote:

From: Chen Qi mailto:qi.c...@windriver.com>>

Add PACKAGECONFIG for autoconf-archive. Default to enable it
to avoid
m4 file conflicts between gnome-common and autoconf-archive.

e.g.
IMAGE_INSTALL_append = " gnome-common autoconf-archive-dev"
bitbake 

We enable autoconf-archive PACKAGECONFIG to avoid the
following error.

Error: Transaction check error:
   file /usr/share/aclocal/ax_check_enable_debug.m4 conflicts
between attempted installs of gnome-common-3.18.0-r0.noarch
and autoconf-archive-dev-2016.09.16-r0.core2_64
   file /usr/share/aclocal/ax_code_coverage.m4 conflicts
between attempted installs of gnome-common-3.18.0-r0.noarch
and autoconf-archive-dev-2016.09.16-r0.core2_64

Signed-off-by: Chen Qi mailto:qi.c...@windriver.com>>
---
  meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
<http://gnome-common_3.18.0.bb> | 4 
  1 file changed, 4 insertions(+)

diff --git
a/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
<http://gnome-common_3.18.0.bb>
b/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
<http://gnome-common_3.18.0.bb>
index 06f3bb3..d874164 100644
---
a/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
<http://gnome-common_3.18.0.bb>
+++
b/meta-oe/recipes-gnome/gnome-common/gnome-common_3.18.0.bb
<http://gnome-common_3.18.0.bb>
@@ -14,6 +14,10 @@ SRC_URI[archive.sha256sum] =
"22569e370ae755e04527b76328befc4c73b62bfd4a572499fd
  EXTRA_AUTORECONF = ""
  DEPENDS = ""
  +# Default to enable autoconf-archive to avoid conflicts
+PACKAGECONFIG ??= "autoconf-archive"
+PACKAGECONFIG[autoconf-archive] = "--with-autoconf-archive,
--without-autoconf-archive, autoconf-archive"
+
  FILES_${PN} += "${datadir}/aclocal"
  FILES_${PN}-dev = ""



-- 
___

Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
<mailto:Openembedded-devel@lists.openembedded.org>
http://lists.openembedded.org/mailman/listinfo/openembedded-devel
<http://lists.openembedded.org/mailman/listinfo/openembedded-devel>




--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-python][PATCH] python-automat: add an initial recipe for version 0.6.0

[ChenQi]

Hi Derek,

I noticed in Automat's setup.py script:

setup_requires=[
'setuptools-scm',
'm2r',
],
install_requires=[
"attrs",
"six",
],

Now the installed_requires entry is handled by the RDEPENDS meta data in 
recipe.

How about the setup_requires entry? Do you know how to handle it in recipe?

I'm asking this because when I build python-automat on my host, I met 
do_compile error because of being able to download the m2r and 
setuptools-scm.


Ideally, we should be able to compile a recipe out without reaching out 
to the Internet, as long as the recipe's tarball is available in the DL_DIR.


Best Regards,
Chen Qi


On 11/02/2017 12:41 AM, Derek Straka wrote:

Signed-off-by: Derek Straka 
---
  meta-python/recipes-devtools/python/python-automat.inc   | 16 
  .../recipes-devtools/python/python-automat_0.6.0.bb  |  2 ++
  .../recipes-devtools/python/python3-automat_0.6.0.bb |  4 
  3 files changed, 22 insertions(+)
  create mode 100644 meta-python/recipes-devtools/python/python-automat.inc
  create mode 100644 meta-python/recipes-devtools/python/python-automat_0.6.0.bb
  create mode 100644 
meta-python/recipes-devtools/python/python3-automat_0.6.0.bb

diff --git a/meta-python/recipes-devtools/python/python-automat.inc 
b/meta-python/recipes-devtools/python/python-automat.inc
new file mode 100644
index 000..1ac5d3b
--- /dev/null
+++ b/meta-python/recipes-devtools/python/python-automat.inc
@@ -0,0 +1,16 @@
+DESCRIPTION = "Self-service finite-state machines for the programmer on the go"
+HOMEPAGE = "https://github.com/glyph/Automat";
+LICENSE = "MIT"
+LIC_FILES_CHKSUM = "file://LICENSE;md5=4ad213bcca81688e94593e5f60c87477"
+
+SRC_URI[md5sum] = "ad7bba58d262d8956d732330cb5ef53d"
+SRC_URI[sha256sum] = 
"3c1fd04ecf08ac87b4dd3feae409542e9bf7827257097b2b6ed5692f69d6f6a8"
+
+PYPI_PACKAGE = "Automat"
+
+inherit pypi
+
+RDEPENDS_${PN} += "\
+   ${PYTHON_PN}-attrs \
+   ${PYTHON_PN}-six \
+"
diff --git a/meta-python/recipes-devtools/python/python-automat_0.6.0.bb 
b/meta-python/recipes-devtools/python/python-automat_0.6.0.bb
new file mode 100644
index 000..8ac8c9c
--- /dev/null
+++ b/meta-python/recipes-devtools/python/python-automat_0.6.0.bb
@@ -0,0 +1,2 @@
+inherit setuptools
+require python-automat.inc
diff --git a/meta-python/recipes-devtools/python/python3-automat_0.6.0.bb 
b/meta-python/recipes-devtools/python/python3-automat_0.6.0.bb
new file mode 100644
index 000..0469e61
--- /dev/null
+++ b/meta-python/recipes-devtools/python/python3-automat_0.6.0.bb
@@ -0,0 +1,4 @@
+inherit setuptools3
+require python-automat.inc
+
+RDEPENDS_${PN} += "${PYTHON_PN}-enum "



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-python][PATCH] python-automat: add an initial recipe for version 0.6.0

[ChenQi]

On 12/22/2017 10:18 PM, Derek Straka wrote:

Chen,

Thanks for bringing this up.  Take a look here to see if this proposed 
patchset fixes your issue: 
http://lists.openembedded.org/pipermail/openembedded-devel/2017-December/116094.html.


-Derek



Derek,

Thanks!
Your patch works.

Best Regards,
Chen Qi

On Fri, Dec 22, 2017 at 4:54 AM, ChenQi <mailto:qi.c...@windriver.com>> wrote:


Hi Derek,

I noticed in Automat's setup.py script:

setup_requires=[
'setuptools-scm',
'm2r',
],
install_requires=[
"attrs",
"six",
],

Now the installed_requires entry is handled by the RDEPENDS meta
data in recipe.
How about the setup_requires entry? Do you know how to handle it
in recipe?

I'm asking this because when I build python-automat on my host, I
met do_compile error because of being able to download the m2r and
setuptools-scm.

Ideally, we should be able to compile a recipe out without
reaching out to the Internet, as long as the recipe's tarball is
available in the DL_DIR.

Best Regards,
Chen Qi


On 11/02/2017 12:41 AM, Derek Straka wrote:

Signed-off-by: Derek Straka mailto:de...@asterius.io>>
---
  meta-python/recipes-devtools/python/python-automat.inc  | 16

  .../recipes-devtools/python/python-automat_0.6.0.bb
<http://python-automat_0.6.0.bb>  | 2 ++
  .../recipes-devtools/python/python3-automat_0.6.0.bb
<http://python3-automat_0.6.0.bb> | 4 
  3 files changed, 22 insertions(+)
  create mode 100644
meta-python/recipes-devtools/python/python-automat.inc
  create mode 100644
meta-python/recipes-devtools/python/python-automat_0.6.0.bb
<http://python-automat_0.6.0.bb>
  create mode 100644
meta-python/recipes-devtools/python/python3-automat_0.6.0.bb
<http://python3-automat_0.6.0.bb>

diff --git
a/meta-python/recipes-devtools/python/python-automat.inc
b/meta-python/recipes-devtools/python/python-automat.inc
new file mode 100644
index 000..1ac5d3b
--- /dev/null
+++ b/meta-python/recipes-devtools/python/python-automat.inc
@@ -0,0 +1,16 @@
+DESCRIPTION = "Self-service finite-state machines for the
programmer on the go"
+HOMEPAGE = "https://github.com/glyph/Automat
<https://github.com/glyph/Automat>"
+LICENSE = "MIT"
+LIC_FILES_CHKSUM =
"file://LICENSE;md5=4ad213bcca81688e94593e5f60c87477"
+
+SRC_URI[md5sum] = "ad7bba58d262d8956d732330cb5ef53d"
+SRC_URI[sha256sum] =
"3c1fd04ecf08ac87b4dd3feae409542e9bf7827257097b2b6ed5692f69d6f6a8"
+
+PYPI_PACKAGE = "Automat"
+
+inherit pypi
+
+RDEPENDS_${PN} += "\
+   ${PYTHON_PN}-attrs \
+   ${PYTHON_PN}-six \
+"
diff --git
a/meta-python/recipes-devtools/python/python-automat_0.6.0.bb
<http://python-automat_0.6.0.bb>
b/meta-python/recipes-devtools/python/python-automat_0.6.0.bb
<http://python-automat_0.6.0.bb>
new file mode 100644
index 000..8ac8c9c
--- /dev/null
+++
b/meta-python/recipes-devtools/python/python-automat_0.6.0.bb
<http://python-automat_0.6.0.bb>
@@ -0,0 +1,2 @@
+inherit setuptools
+require python-automat.inc
diff --git
a/meta-python/recipes-devtools/python/python3-automat_0.6.0.bb
<http://python3-automat_0.6.0.bb>
b/meta-python/recipes-devtools/python/python3-automat_0.6.0.bb
<http://python3-automat_0.6.0.bb>
new file mode 100644
index 000..0469e61
--- /dev/null
+++
b/meta-python/recipes-devtools/python/python3-automat_0.6.0.bb
<http://python3-automat_0.6.0.bb>
@@ -0,0 +1,4 @@
+inherit setuptools3
+require python-automat.inc
+
+RDEPENDS_${PN} += "${PYTHON_PN}-enum "





--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] Dependency loop in cryptsetup

[ChenQi]

On 02/28/2018 11:48 AM, Md Ismail, Zakiah wrote:

Hi,

I want to enable cyrptsetup in systemd by adding PACKAGECONFIG_append_pn-systemd = " 
cryptsetup" in my local.conf; however, I encounter the dependency loop issue. The 
issue is similar to this: https://bugzilla.yoctoproject.org/show_bug.cgi?id=11856. The 
difference is I'm using the latest commit from meta-openembedded rocko branch. Please 
advise on the steps to solve this issue. Thanks in advance!

Regards,
Zakiah



I'd like to provide some information here, as I worked on this issue.
When I first encountered this circular dependency problem, I solved this 
problem by separating lvm2 into two recipes and making them having 
different dependencies.
After that, I found that if we do so, then using tools from lvm2 would 
have problem. I did some investigation and found that lvm2 and 
libdevmapper recipes must have the same dependencies to work correctly 
at runtime. So I sent patches to fix this runtime problem, and 
introducing this circular dependency problem back.


That's the whole history of this problem. The current status is that, we 
still have this circular dependency problem.


I thought about this problem for quite a while at that. The circular 
dependency is about the following three packages.

libdevmapper, udev, cryptsetup.
cryptsetup depends on libdevmapper, I checked the its configure and can 
confirm that it's a necessity.
libdevmapper depends on udev (provided by systemd), and this dependency 
could be broke by removing udev from PACKAGECONFIG, with the cost of 
losing of functionality.
udev depends on cryptsetup, and as this is what user wants, the 
dependency could not be broke.


In summary, your only option at the moment is to remove 'udev' from lvm2 
and libdevmapper with the cost of losing some functionality.


P.S.
To totally solve this problem, in the scheme of OE, the only option 
seems to break systemd into multiple recipes, which I really don't 
suggest to do.


Best Regards,
Chen Qi
--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-oe][PATCH 5/5] gperftools: Enable on mips

[ChenQi]

On 10/03/2017 12:06 AM, Khem Raj wrote:

Signed-off-by: Khem Raj 
---
  meta-oe/recipes-support/gperftools/gperftools_2.6.1.bb | 4 
  1 file changed, 4 deletions(-)

diff --git a/meta-oe/recipes-support/gperftools/gperftools_2.6.1.bb 
b/meta-oe/recipes-support/gperftools/gperftools_2.6.1.bb
index 23284670b..8ce376c91 100644
--- a/meta-oe/recipes-support/gperftools/gperftools_2.6.1.bb
+++ b/meta-oe/recipes-support/gperftools/gperftools_2.6.1.bb
@@ -18,10 +18,6 @@ inherit autotools
  
  S = "${WORKDIR}/git"
  
-# On mips, we have the following error.

-#   do_page_fault(): sending SIGSEGV to ls for invalid read access from 
0008
-#   Segmentation fault (core dumped)
-COMPATIBLE_HOST = "(i.86|x86_64|powerpc|powerpc64|arm|aarch64).*-linux*"
  # Disable thumb1
  # {standard input}: Assembler messages:
  # {standard input}:434: Error: lo register required -- `ldr pc,[sp]'


Hi Khem,

On qemumips, I still get the error.

   root@qemumips:~# LD_PRELOAD=libtcmalloc.so ls /
[  407.636947] do_page_fault(): sending SIGSEGV to ls for 
invalid read access from 0008
[  407.646240] epc = 77e1ae08 in 
libunwind.so.8.0.1[77e18000+c000]
[  407.651497] ra  = 77e1b3ac in 
libunwind.so.8.0.1[77e18000+c000]

Segmentation fault (core dumped)

The libc is glibc. I didn't test for musl.

I guess you've tested mips for musl. So how about we add this setting 
back but limit it to glibc?

Something like:
COMPATIBLE_HOST_libc-glibc = 
"(i.86|x86_64|powerpc|powerpc64|arm|aarch64).*-linux*"


Best Regards,
Chen Qi
--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [OE-core] [PATCH 1/2] systemd: upgrade to 239

[ChenQi]

Hi All,

In brief, I also agree with Martin and will send out patch to add back 
these alternatives for systemd.

Just one thing to confirm.
Martin, you have a customized busybox defconfig which enables init 
utilties by default, right?


Below are more details and history.

1) Why removing alternatives of init utilities from systemd?
I removed the alternatives for init utilities from systemd because I 
thought these init utilities are basically bound to init manager. In 
other words, there's no point to use reboot from busybox when init 
manager is systemd. Same logic applies to other utilities. So I removed 
these alternatives of the init utilties from systemd, and also moved 
related busybox configs to init.cfg.


2) Why reboot is moved out of busybox init.cfg and added back as an 
alternative for systemd?
The above logic misses one thing, the live image. In oe-core, the 
initramfs for the live image is implemented as a bunch of shell scripts. 
And they need the 'reboot' command. More particularly, they call 'reboot 
-f'. That means they only want kernel to do reboot directly, discarding 
PID 1.
There's a yocto bug for this. 
https://bugzilla.yoctoproject.org/show_bug.cgi?id=12914


3) The current situation and potential problems
For oe-core, currently we have:
root@qemux86-64:~# ls -l /sbin/reboot
lrwxrwxrwx 1 root root 20 Oct 18 10:50 /sbin/reboot -> /sbin/reboot.systemd
root@qemux86-64:~# ls -l /sbin/reboot.systemd
lrwxrwxrwx 1 root root 16 Oct 18 10:11 /sbin/reboot.systemd -> 
../bin/systemctl

So for oe-core's default configs, things work well.

BUT, I'm now somewhat worried about users who have customized busybox 
defconfig which enables init utilities by default. The current oe-core 
release will cause regression for them if they use systemd +  busybox.


4) What I will do.
I'll add back these init utilities for systemd to avoid regression.

Best Regards,
Chen Qi

On 10/18/2018 06:55 PM, Martin Jansa wrote:
Using u-a for the systemd symlinks to systemctl seems like best 
solution to me, I've already suggested it in:

http://lists.openembedded.org/pipermail/openembedded-core/2018-October/156597.html

On Thu, Oct 18, 2018 at 12:33 PM nick83ola > wrote:


Another thing:

previously
ALTERNATIVE_PRIORITY[reboot] = "100"
was
ALTERNATIVE_PRIORITY[reboot] = "300"

diff of the relevant part:

539,566c559
< # TODO:
< # u-a for runlevel and telinit
<
< ALTERNATIVE_${PN} = "init halt reboot shutdown poweroff runlevel
resolv-conf"
<
< ALTERNATIVE_TARGET[init] = "${rootlibexecdir}/systemd/systemd"
< ALTERNATIVE_LINK_NAME[init] = "${base_sbindir}/init"
< ALTERNATIVE_PRIORITY[init] ?= "300"
<
< ALTERNATIVE_TARGET[halt] = "${base_bindir}/systemctl"
< ALTERNATIVE_LINK_NAME[halt] = "${base_sbindir}/halt"
< ALTERNATIVE_PRIORITY[halt] ?= "300"
<
< ALTERNATIVE_TARGET[reboot] = "${base_bindir}/systemctl"
< ALTERNATIVE_LINK_NAME[reboot] = "${base_sbindir}/reboot"
< ALTERNATIVE_PRIORITY[reboot] ?= "300"
<
< ALTERNATIVE_TARGET[shutdown] = "${base_bindir}/systemctl"
< ALTERNATIVE_LINK_NAME[shutdown] = "${base_sbindir}/shutdown"
< ALTERNATIVE_PRIORITY[shutdown] ?= "300"
<
< ALTERNATIVE_TARGET[poweroff] = "${base_bindir}/systemctl"
< ALTERNATIVE_LINK_NAME[poweroff] = "${base_sbindir}/poweroff"
< ALTERNATIVE_PRIORITY[poweroff] ?= "300"
<
< ALTERNATIVE_TARGET[runlevel] = "${base_bindir}/systemctl"
< ALTERNATIVE_LINK_NAME[runlevel] = "${base_sbindir}/runlevel"
< ALTERNATIVE_PRIORITY[runlevel] ?= "300"
---
> ALTERNATIVE_${PN} = "resolv-conf reboot"
570a564,566
>
> ALTERNATIVE_LINK_NAME[reboot] = "${base_sbindir}/reboot"
> ALTERNATIVE_PRIORITY[reboot] = "100"

On Thu, 18 Oct 2018 at 11:29, nick83ola mailto:nick83...@gmail.com>> wrote:

> Hi Chen,
>
> regarding this issue:
>
> >On 07/26/2018 07:05 AM, Ricardo Salveti wrote:
> >> On Mon, Jul 16, 2018 at 11:05 PM, Chen Qi http://windriver.com>> wrote:
> >>> Upgrade systemd to 239.
> >>>
> >>> 3. update-alternatives changes
> >>>The utilities like shutdown, poweroff, etc. are now
created as symlinks
> >>>at do_install. So there's no need to use
update-alternatives mechanism
> >>>anymore to create the symlinks now. In addtion, I don't
think we now
> >>>support multiple init systems at one running system, so
there's really
> >>>no need to use update-alternatives mechanism here.
> >> I noticed that reboot stopped working locally as I had busybox
> >> installed together with systemd, and the busybox version ended up
> >> being used as it was provided via update-alternatives.
> >>
> >> Looking for possible similar broken links, I found that
> >> update-alternatives ended up pointing reboot, halt and
poweroff to the
> >> busybox binary instead of systemctl. Should we rev

Re: [oe] [meta-networking][PATCH 1/2] netkit-rsh: add tag to CVE patch

[ChenQi]

Sorry, please ignore this patch. The CVE ID is wrong. V2 is coming.

Best Regards,
Chen Qi

On 05/07/2019 10:11 AM, Chen Qi wrote:

Signed-off-by: Chen Qi 
---
  .../netkit-rsh/netkit-rsh/CVE-2019-7282-and-CVE-2019-7283.patch | 2 ++
  1 file changed, 2 insertions(+)

diff --git 
a/meta-networking/recipes-netkit/netkit-rsh/netkit-rsh/CVE-2019-7282-and-CVE-2019-7283.patch
 
b/meta-networking/recipes-netkit/netkit-rsh/netkit-rsh/CVE-2019-7282-and-CVE-2019-7283.patch
index 4381f5b..dc71141 100644
--- 
a/meta-networking/recipes-netkit/netkit-rsh/netkit-rsh/CVE-2019-7282-and-CVE-2019-7283.patch
+++ 
b/meta-networking/recipes-netkit/netkit-rsh/netkit-rsh/CVE-2019-7282-and-CVE-2019-7283.patch
@@ -10,6 +10,8 @@ Origin: 
https://github.com/openssh/openssh-portable/commit/6010c0303a422a9c5fa88
  Upstream-Status: Backport [Debian]
  
[https://sources.debian.org/src/netkit-rsh/0.17-20/debian/patches/fix-CVE-2018-20685-and-CVE-2019-6111.patch]
  
+CVE: CVE-2018-7282 CVE-2018-7283

+
  Signed-off-by: Yi Zhao 
  ---
   rcp/rcp.c | 5 +



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel

Re: [oe] [meta-networking][PATCH] postfix: fix postinstall to avoid do_rootfs failure

[ChenQi]

Please ignore this patch.
I'll send out a new solution.

Regards,
Chen Qi

On 07/15/2019 01:09 PM, Chen Qi wrote:

When lsb is enabled, that is, 'linuxstdbase' is in OVERRIDES,
do_rootfs would fail if both 'msmtp' and 'postfix' are installed.
This is because they both try to create /usr/lib/sendmail link if
it does not exist. So we need to fix the check to avoid non-zero
exit status.

Signed-off-by: Chen Qi 
---
  meta-networking/recipes-daemons/postfix/postfix.inc | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/meta-networking/recipes-daemons/postfix/postfix.inc 
b/meta-networking/recipes-daemons/postfix/postfix.inc
index 3d4f1df..021129c 100644
--- a/meta-networking/recipes-daemons/postfix/postfix.inc
+++ b/meta-networking/recipes-daemons/postfix/postfix.inc
@@ -221,7 +221,7 @@ pkg_postinst_${PN} () {
  else
  if ${@'true' if 'linuxstdbase' in d.getVar('DISTROOVERRIDES', False) 
else 'false'}; then
  # /usr/lib/sendmail is required by LSB core test
-[ ! -L $D/usr/lib/sendmail ] && ln -sf ${sbindir}/sendmail 
$D/usr/lib/
+[ ! -L $D/usr/lib/sendmail ] && ln -sf ${sbindir}/sendmail 
$D/usr/lib/ || true
  fi
  
  touch $D/etc/aliases



--
___
Openembedded-devel mailing list
Openembedded-devel@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-devel