[OE-core] [PATCH 1/2] Images: handle rowstride != width*bytes_per_pixel

2014-10-07 Thread Olaf Mandel 
From 0793aa7daf4b594fae1b412ab16a7d42d8c8560b Mon Sep 17 00:00:00 2001
From: Olaf Mandel o.man...@menlosystems.com
Date: Tue, 7 Oct 2014 15:12:08 +0200
Subject: [PATCH 1/2] Images: handle rowstride != width*bytes_per_pixel

Up till now, image data with a rowstride larger than the value
expected from width and bytes_per_pixel caused a trapezoid
distortion of the displayed image.

Signed-off-by: Olaf Mandel o.man...@menlosystems.com
---
 ChangeLog|7 +++
 psplash-fb.c |   13 +++--
 psplash-fb.h |1 +
 psplash.c|2 ++
 4 files changed, 17 insertions(+), 6 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 8cf2156..f4fd0ec 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,10 @@
+2014-10-07  Olaf Mandel o.man...@menlosystems.com
+
+   * psplash-fb.c:
+   * psplash-fb.h:
+   * psplash.c:
+   Images: handle rowstride != width*bytes_per_pixel
+
 2009-05-28  Richard Purdie rpur...@linux.intel.com
 
* psplash-fb.c:
diff --git a/psplash-fb.c b/psplash-fb.c
index 71740cd..bd9cd9d 100644
--- a/psplash-fb.c
+++ b/psplash-fb.c
@@ -370,13 +370,14 @@ psplash_fb_draw_image (PSplashFB*fb,
   int  img_width,
   int  img_height,
   int  img_bytes_per_pixel,
+  int  img_rowstride,
   uint8   *rle_data)
 {
   uint8   *p = rle_data;
   int  dx = 0, dy = 0,  total_len;
   unsigned int len;
 
-  total_len = img_width * img_height * img_bytes_per_pixel;
+  total_len = img_rowstride * img_height;
 
   /* FIXME: Optimise, check for over runs ... */
   while ((p - rle_data)  total_len)
@@ -391,11 +392,11 @@ psplash_fb_draw_image (PSplashFB*fb,
 
  do
{
- if (img_bytes_per_pixel  4 || *(p+3))
+ if ((img_bytes_per_pixel  4 || *(p+3))  dx  img_width)
psplash_fb_plot_pixel (fb, x+dx, y+dy, *(p), *(p+1), *(p+2));
- if (++dx = img_width) { dx=0; dy++; }
+ if (++dx * img_bytes_per_pixel = img_rowstride) { dx=0; dy++; }
}
- while (--len  (p - rle_data)  total_len);
+ while (--len);
 
  p += img_bytes_per_pixel;
}
@@ -405,9 +406,9 @@ psplash_fb_draw_image (PSplashFB*fb,
 
  do
{
- if (img_bytes_per_pixel  4 || *(p+3))
+ if ((img_bytes_per_pixel  4 || *(p+3))  dx  img_width)
psplash_fb_plot_pixel (fb, x+dx, y+dy, *(p), *(p+1), *(p+2));
- if (++dx = img_width) { dx=0; dy++; }
+ if (++dx * img_bytes_per_pixel = img_rowstride) { dx=0; dy++; }
  p += img_bytes_per_pixel;
}
  while (--len  (p - rle_data)  total_len);
diff --git a/psplash-fb.h b/psplash-fb.h
index ef5b39e..42592ed 100644
--- a/psplash-fb.h
+++ b/psplash-fb.h
@@ -82,6 +82,7 @@ psplash_fb_draw_image (PSplashFB*fb,
   int  img_width, 
   int  img_height,
   int  img_bytes_pre_pixel,
+  int  img_rowstride,
   uint8   *rle_data);
 
 void
diff --git a/psplash.c b/psplash.c
index 09cf0d0..543f67e 100644
--- a/psplash.c
+++ b/psplash.c
@@ -274,6 +274,7 @@ main (int argc, char** argv)
 POKY_IMG_WIDTH,
 POKY_IMG_HEIGHT,
 POKY_IMG_BYTES_PER_PIXEL,
+POKY_IMG_ROWSTRIDE,
 POKY_IMG_RLE_PIXEL_DATA);
 
   /* Draw progress bar border */
@@ -283,6 +284,7 @@ main (int argc, char** argv)
 BAR_IMG_WIDTH,
 BAR_IMG_HEIGHT,
 BAR_IMG_BYTES_PER_PIXEL,
+BAR_IMG_ROWSTRIDE,
 BAR_IMG_RLE_PIXEL_DATA);
 
   psplash_draw_progress (fb, 0);
-- 
1.7.10.4




signature.asc
Description: OpenPGP digital signature
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

Re: [OE-core] [PATCH 1/2] Images: handle rowstride != width*bytes_per_pixel

2014-10-07 Thread Olaf Mandel 
Am 07.10.2014 18:43, schrieb Olaf Mandel:
 Up till now, image data with a rowstride larger than the value
 expected from width and bytes_per_pixel caused a trapezoid
 distortion of the displayed image.
 
 Signed-off-by: Olaf Mandel o.man...@menlosystems.com
-Snipp-

Hi,

my apologies: I should have said that this patch and the next are
against the psplash repository:

http://git.yoctoproject.org/cgit/cgit.cgi/psplash

This is why I CCed the pslash maintainer.

Best regards,
Olaf Mandel



signature.asc
Description: OpenPGP digital signature
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH 1/1] Remove $COREBASE/LICENSE from LIC_FILES_CHKSUM

2016-10-19 Thread Olaf Mandel 
Several recipes reference the LICENSE file in their LIC_FILES_CHKSUM
variable as ${COREBASE}/LICENSE. This forces distribution providers to
keep this file verbatim or to overload the affected recipes. The section
"Moving to the Yocto Project 1.6 Release" in the Yocto manual suggests
removing the LICENSE file where possible.

Remove LICENSE in cases where COPYING.MIT is also given and replace
LICENSE with COPYING.MIT if the former was the only entry. All modified
recipes specify LICENSE = "MIT" and none of the in-tree files specify a
different license either.

As the packages do not change (the license files are not contained in
them), do not increase PR.
---
 meta-skeleton/recipes-skeleton/useradd/useradd-example.bb | 3 +--
 meta/recipes-bsp/formfactor/formfactor_0.0.bb | 3 +--
 meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb| 3 +--
 meta/recipes-core/systemd/systemd-compat-units.bb | 2 +-
 meta/recipes-core/systemd/systemd-systemctl-native.bb | 2 +-
 meta/recipes-core/udev/udev-extraconf_1.1.bb  | 3 +--
 meta/recipes-devtools/devel-config/distcc-config.bb   | 3 +--
 meta/recipes-devtools/devel-config/nfs-export-root.bb | 3 +--
 meta/recipes-devtools/run-postinsts/run-postinsts_1.0.bb  | 3 +--
 meta/recipes-graphics/wayland/weston-init.bb  | 2 +-
 meta/recipes-sato/shutdown-desktop/shutdown-desktop.bb| 3 +--
 11 files changed, 11 insertions(+), 19 deletions(-)

diff --git a/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb 
b/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb
index d3c02d5..20a101a 100644
--- a/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb
+++ b/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb
@@ -3,8 +3,7 @@ DESCRIPTION = "This recipe serves as an example for using 
features from useradd.
 SECTION = "examples"
 PR = "r1"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 SRC_URI = "file://file1 \
file://file2 \
diff --git a/meta/recipes-bsp/formfactor/formfactor_0.0.bb 
b/meta/recipes-bsp/formfactor/formfactor_0.0.bb
index 5c30bd5..53cf1cf 100644
--- a/meta/recipes-bsp/formfactor/formfactor_0.0.bb
+++ b/meta/recipes-bsp/formfactor/formfactor_0.0.bb
@@ -1,8 +1,7 @@
 SUMMARY = "Device formfactor information"
 SECTION = "base"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 PR = "r45"
 
 SRC_URI = "file://config file://machconfig"
diff --git a/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb 
b/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb
index 51a76b4..b5f6895 100644
--- a/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb
+++ b/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb
@@ -4,8 +4,7 @@ DEPENDS = "ppp"
 RDEPENDS_${PN} = "ppp"
 PR = "r8"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 SRC_URI = "file://host-peer \
file://ppp-dialin"
diff --git a/meta/recipes-core/systemd/systemd-compat-units.bb 
b/meta/recipes-core/systemd/systemd-compat-units.bb
index 421fc06..aa20014 100644
--- a/meta/recipes-core/systemd/systemd-compat-units.bb
+++ b/meta/recipes-core/systemd/systemd-compat-units.bb
@@ -1,7 +1,7 @@
 SUMMARY = "Enhances systemd compatilibity with existing SysVinit scripts"
 
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 PR = "r29"
 
diff --git a/meta/recipes-core/systemd/systemd-systemctl-native.bb 
b/meta/recipes-core/systemd/systemd-systemctl-native.bb
index fbdc9c0..fadc843 100644
--- a/meta/recipes-core/systemd/systemd-systemctl-native.bb
+++ b/meta/recipes-core/systemd/systemd-systemctl-native.bb
@@ -1,7 +1,7 @@
 SUMMARY = "Wrapper for enabling systemd services"
 
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 PR = "r6"
 
diff --git a/meta/recipes-core/udev/udev-extraconf_1.1.bb 
b/meta/recipes-core/udev/udev-extraconf_1.1.bb
index ecd4a8a..ae12550 100644
--- a/meta/recipes-core/udev/udev-extraconf_1.1.bb
+++ b/meta/recipes-core/udev/udev-extraconf_1.1.bb
@@ -1,8 +1,7 @@
 SUMMARY

[OE-core] [PATCH 0/1] $COREBASE/LICENSE in LIC_FILES_CHKSUM

2016-10-19 Thread Olaf Mandel 
I ran into QA issues when inadvertently replacing OEs LICENSE with
Yoctos LICENSE file. Why should recipes depend on the license for
the complete collection of layers? At least for "normal" recipes
that generate target device packages, this makes no sense to me.

I patched out the global LICENSE file from all those recipes that
are short enough and where I feel reasonably confident of not
having actually changed the claimed license terms in any way.

There are additional recipes that reference the global LICENSE
file, but where I am not sure if a less blanket license declaration
is appropriate or not. Especially for meta-toolchain and
build-appliance-image, the global LICENSE is probably correct.
Any comments?

If there are really recipes where the global LICENSE file is
appropriate, how to allow people running their own combo-layer to
quickly replace the file? Maybe add a variable CORELICENSE_MD5 that
gets used by all such recipes?

-- 
2.1.4

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

Re: [OE-core] [PATCH 1/1] Remove $COREBASE/LICENSE from LIC_FILES_CHKSUM

2016-10-20 Thread Olaf Mandel 
Hello,

Am 20.10.2016 um 12:44 schrieb Peter Kjellerstedt:
>> -Original Message-
-Snipp-
>> Subject: [OE-core] [PATCH 1/1] Remove $COREBASE/LICENSE from
-Snipp-
> 
> Even after this is applied, there are still nine recipes in OE-Core 
> that reference ${COREBASE}/LICENSE. There is also an example in 
> dev-manual-common-tasks.xml with a reference to ${COREBASE}/LICENSE. 
> AFAICT, they all look just like the ones that this patch removes. 
> Any reason they were not fixed as well?
> 
"Why not patch the nine remaining recipes": see my cover letter
"[OE-core][PATCH 0/1] $COREBASE/LICENSE in LIC_FILES_CHKSUM":

>>> I patched out the global LICENSE file from all those recipes that
>>> are short enough and where I feel reasonably confident of not
>>> having actually changed the claimed license terms in any way.
>>> 
>>> There are additional recipes that reference the global LICENSE
>>> file, but where I am not sure if a less blanket license declaration
>>> is appropriate or not. Especially for meta-toolchain and
>>> build-appliance-image, the global LICENSE is probably correct.
>>> Any comments?

So please comment away  :-)

> There are also 17 recipes in OpenEmbedded with a reference to 
> ${COREBASE}/LICENSE. They too need to be fixed.
> 
Same as for dev-manual-common-tasks.xml: I was not a aware of those and
not all belong on the openembedded-core list. And for the example: it
should only be patched once this patch for formfactor_0.0.bb is accepted.

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

[OE-core] [PATCH] combo-layer: handle ambiguous git arguments

2016-10-17 Thread Olaf Mandel 
If a branch/src-repository has the same name as a file/directory, git
since 1.4.0(?) gives an error like the one below:
ambiguous argument 'bitbake': both revision and filename
Use '--' to separate paths from revisions, like this:
'git  [...] -- [...]'

Add two dashes to make the intent clear.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 scripts/combo-layer | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/combo-layer b/scripts/combo-layer
index 4a210fb..089b65f 100755
--- a/scripts/combo-layer
+++ b/scripts/combo-layer
@@ -426,7 +426,7 @@ file_exclude = %s''' % (name, file_filter or '', 
repo.get('file_exclude',
 merge.append(name)
 # Root all commits which have no parent in the common
 # ancestor in the new repository.
-for start in runcmd('git log --pretty=format:%%H 
--max-parents=0 %s' % name).split('\n'):
+for start in runcmd('git log --pretty=format:%%H 
--max-parents=0 %s --' % name).split('\n'):
 runcmd('git replace --graft %s %s' % (start, startrev))
 try:
 runcmd(merge)
-- 
2.1.4

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

[OE-core] [PATCH] combo-layer: fix crashes on wrong tempfile usage

2016-10-17 Thread Olaf Mandel 
When calling tempfile.NamedTemporaryFile().write(str()), at least on
Python 3.4.2 this fails with this error:
TypeError: 'str' does not support the buffer interface

Change the file-mode for all such files from binary to text mode.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 scripts/combo-layer | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/scripts/combo-layer b/scripts/combo-layer
index b90bfc8..4a210fb 100755
--- a/scripts/combo-layer
+++ b/scripts/combo-layer
@@ -329,7 +329,7 @@ def action_init(conf, args):
 # one. The commit should be in both repos with
 # the same tree, but better check here.
 tree = runcmd('git show -s --pretty=format:%%T %s' % 
rev).strip()
-with tempfile.NamedTemporaryFile() as editor:
+with tempfile.NamedTemporaryFile(mode='wt') as editor:
 editor.write('''cat >$1 <http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH v2] texi2html: Allow compiling out-of-source

2016-10-21 Thread Olaf Mandel 
Compiling texi2html 5.0 out-of-source with USE_NLS set to no failed
because it tried to copy from srcdir without using that variable.
Fix this issue and add a reference to the upstream commit.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 .../files/0001-Allow-compiling-out-of-source.patch | 39 ++
 meta/recipes-extended/texi2html/texi2html_5.0.bb   |  1 +
 2 files changed, 40 insertions(+)
 create mode 100644 
meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch

diff --git 
a/meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch
 
b/meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch
new file mode 100644
index 000..0cf025f
--- /dev/null
+++ 
b/meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch
@@ -0,0 +1,39 @@
+From: Olaf Mandel <o.man...@menlosystems.com>
+Date: Fri, 21 Oct 2016 13:04:44 +
+Subject: [PATCH] Allow compiling out-of-source
+
+Upstream-Status: Backport of [svn://svn.sv.gnu.org/texinfo/trunk r3602]
+---
+ Makefile.am | 2 +-
+ Makefile.in | 2 +-
+ 2 files changed, 2 insertions(+), 2 deletions(-)
+
+diff --git a/Makefile.am b/Makefile.am
+index 3447463..c9b5b5c 100644
+--- a/Makefile.am
 b/Makefile.am
+@@ -208,7 +208,7 @@ i18n/en.thl i18n/: 
$(po_document_dir)/po_document/$(PACKAGE)_document.pot
+ done; \
+ msgexec -i $< "$(srcdir)/gettext_to_separated.pl" | 
"$(srcdir)/separated_to_hash.pl" en > i18n/en.thl; \
+   else \
+-cp -p i18n_ref/*.thl i18n; \
++cp -p "$(srcdir)/i18n_ref/"*.thl i18n; \
+   fi
+ 
+ i18n_ref:
+diff --git a/Makefile.in b/Makefile.in
+index 4264b37..a13f84d 100644
+--- a/Makefile.in
 b/Makefile.in
+@@ -1126,7 +1126,7 @@ i18n/en.thl i18n/: 
$(po_document_dir)/po_document/$(PACKAGE)_document.pot
+ done; \
+ msgexec -i $< "$(srcdir)/gettext_to_separated.pl" | 
"$(srcdir)/separated_to_hash.pl" en > i18n/en.thl; \
+   else \
+-cp -p i18n_ref/*.thl i18n; \
++cp -p "$(srcdir)/i18n_ref/"*.thl i18n; \
+   fi
+ 
+ i18n_ref:
+-- 
+2.1.4
+
diff --git a/meta/recipes-extended/texi2html/texi2html_5.0.bb 
b/meta/recipes-extended/texi2html/texi2html_5.0.bb
index eac289e..ae64816 100644
--- a/meta/recipes-extended/texi2html/texi2html_5.0.bb
+++ b/meta/recipes-extended/texi2html/texi2html_5.0.bb
@@ -7,6 +7,7 @@ LIC_FILES_CHKSUM = 
"file://COPYING;md5=59530bdf33659b29e73d4adb9f9f6552"
 PR = "r2"
 
 SRC_URI = "${SAVANNAH_GNU_MIRROR}/texi2html/${BPN}-${PV}.tar.bz2 \
+   file://0001-Allow-compiling-out-of-source.patch \
"
 
 SRC_URI[md5sum] = "f15ac876fcdc8be865b16535f480aa54"
-- 
2.1.4

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

[OE-core] Forgot Signed-off-by lines

2016-10-21 Thread Olaf Mandel 
I forgot to signoff two of my last patches, so here is a v2 of both
of them.

Olaf Mandel (2):
  combo-layer: handle file_exclude matching dirs
  Remove $COREBASE/LICENSE from LIC_FILES_CHKSUM

-- 
2.1.4

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

[OE-core] [PATCH v2] combo-layer: handle file_exclude matching dirs

2016-10-21 Thread Olaf Mandel 
If file_exclude matches a directory, os.unlink() got called with this
directory as an argument.

Filter out paths that end in a directory separator.

This still leaves the (then empty) directories, but this does not affect
the git commit.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 scripts/combo-layer | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/scripts/combo-layer b/scripts/combo-layer
index 089b65f..d04d88b 100755
--- a/scripts/combo-layer
+++ b/scripts/combo-layer
@@ -294,6 +294,8 @@ def action_init(conf, args):
 # again. Uses the list of files created by tar (easier
 # than walking the tree).
 for file in files.split('\n'):
+if file.endswith(os.path.sep):
+continue
 for pattern in exclude_patterns:
 if fnmatch.fnmatch(file, pattern):
 os.unlink(os.path.join(*([extract_dir] + 
['..'] * subdir_components + [file])))
-- 
2.1.4

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

[OE-core] [PATCH v2] Remove $COREBASE/LICENSE from LIC_FILES_CHKSUM

2016-10-21 Thread Olaf Mandel 
Several recipes reference the LICENSE file in their LIC_FILES_CHKSUM
variable as ${COREBASE}/LICENSE. This forces distribution providers to
keep this file verbatim or to overload the affected recipes. The section
"Moving to the Yocto Project 1.6 Release" in the Yocto manual suggests
removing the LICENSE file where possible.

Remove LICENSE in cases where COPYING.MIT is also given and replace
LICENSE with COPYING.MIT if the former was the only entry. All modified
recipes specify LICENSE = "MIT" and none of the in-tree files specify a
different license either.

As the packages do not change (the license files are not contained in
them), do not increase PR.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 meta-skeleton/recipes-skeleton/useradd/useradd-example.bb | 3 +--
 meta/recipes-bsp/formfactor/formfactor_0.0.bb | 3 +--
 meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb| 3 +--
 meta/recipes-core/systemd/systemd-compat-units.bb | 2 +-
 meta/recipes-core/systemd/systemd-systemctl-native.bb | 2 +-
 meta/recipes-core/udev/udev-extraconf_1.1.bb  | 3 +--
 meta/recipes-devtools/devel-config/distcc-config.bb   | 3 +--
 meta/recipes-devtools/devel-config/nfs-export-root.bb | 3 +--
 meta/recipes-devtools/run-postinsts/run-postinsts_1.0.bb  | 3 +--
 meta/recipes-graphics/wayland/weston-init.bb  | 2 +-
 meta/recipes-sato/shutdown-desktop/shutdown-desktop.bb| 3 +--
 11 files changed, 11 insertions(+), 19 deletions(-)

diff --git a/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb 
b/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb
index d3c02d5..20a101a 100644
--- a/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb
+++ b/meta-skeleton/recipes-skeleton/useradd/useradd-example.bb
@@ -3,8 +3,7 @@ DESCRIPTION = "This recipe serves as an example for using 
features from useradd.
 SECTION = "examples"
 PR = "r1"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 SRC_URI = "file://file1 \
file://file2 \
diff --git a/meta/recipes-bsp/formfactor/formfactor_0.0.bb 
b/meta/recipes-bsp/formfactor/formfactor_0.0.bb
index 5c30bd5..53cf1cf 100644
--- a/meta/recipes-bsp/formfactor/formfactor_0.0.bb
+++ b/meta/recipes-bsp/formfactor/formfactor_0.0.bb
@@ -1,8 +1,7 @@
 SUMMARY = "Device formfactor information"
 SECTION = "base"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 PR = "r45"
 
 SRC_URI = "file://config file://machconfig"
diff --git a/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb 
b/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb
index 51a76b4..b5f6895 100644
--- a/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb
+++ b/meta/recipes-connectivity/ppp-dialin/ppp-dialin_0.1.bb
@@ -4,8 +4,7 @@ DEPENDS = "ppp"
 RDEPENDS_${PN} = "ppp"
 PR = "r8"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 SRC_URI = "file://host-peer \
file://ppp-dialin"
diff --git a/meta/recipes-core/systemd/systemd-compat-units.bb 
b/meta/recipes-core/systemd/systemd-compat-units.bb
index 421fc06..aa20014 100644
--- a/meta/recipes-core/systemd/systemd-compat-units.bb
+++ b/meta/recipes-core/systemd/systemd-compat-units.bb
@@ -1,7 +1,7 @@
 SUMMARY = "Enhances systemd compatilibity with existing SysVinit scripts"
 
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690"
+LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 PR = "r29"
 
diff --git a/meta/recipes-core/systemd/systemd-systemctl-native.bb 
b/meta/recipes-core/systemd/systemd-systemctl-native.bb
index fbdc9c0..fadc843 100644
--- a/meta/recipes-core/systemd/systemd-systemctl-native.bb
+++ b/meta/recipes-core/systemd/systemd-systemctl-native.bb
@@ -1,7 +1,7 @@
 SUMMARY = "Wrapper for enabling systemd services"
 
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690"
+LIC_FILES

[OE-core] [PATCH] texi2html: Allow compiling out-of-source

2016-10-21 Thread Olaf Mandel 
Compiling texi2html 5.0 out-of-source with USE_NLS set to no failed
because it trieed to copy from srcdir without using that variable.
Fix this issue.

No need to send upstream: current trunk already contains this.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 .../files/0001-Allow-compiling-out-of-source.patch | 38 ++
 meta/recipes-extended/texi2html/texi2html_5.0.bb   |  1 +
 2 files changed, 39 insertions(+)
 create mode 100644 
meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch

diff --git 
a/meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch
 
b/meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch
new file mode 100644
index 000..754f0a2
--- /dev/null
+++ 
b/meta/recipes-extended/texi2html/files/0001-Allow-compiling-out-of-source.patch
@@ -0,0 +1,38 @@
+From: Olaf Mandel <o.man...@menlosystems.com>
+Date: Fri, 21 Oct 2016 13:04:44 +
+Subject: [PATCH] Allow compiling out-of-source
+
+---
+ Makefile.am | 2 +-
+ Makefile.in | 2 +-
+ 2 files changed, 2 insertions(+), 2 deletions(-)
+
+diff --git a/Makefile.am b/Makefile.am
+index 3447463..c9b5b5c 100644
+--- a/Makefile.am
 b/Makefile.am
+@@ -208,7 +208,7 @@ i18n/en.thl i18n/: 
$(po_document_dir)/po_document/$(PACKAGE)_document.pot
+ done; \
+ msgexec -i $< "$(srcdir)/gettext_to_separated.pl" | 
"$(srcdir)/separated_to_hash.pl" en > i18n/en.thl; \
+   else \
+-cp -p i18n_ref/*.thl i18n; \
++cp -p "$(srcdir)/i18n_ref/"*.thl i18n; \
+   fi
+ 
+ i18n_ref:
+diff --git a/Makefile.in b/Makefile.in
+index 4264b37..a13f84d 100644
+--- a/Makefile.in
 b/Makefile.in
+@@ -1126,7 +1126,7 @@ i18n/en.thl i18n/: 
$(po_document_dir)/po_document/$(PACKAGE)_document.pot
+ done; \
+ msgexec -i $< "$(srcdir)/gettext_to_separated.pl" | 
"$(srcdir)/separated_to_hash.pl" en > i18n/en.thl; \
+   else \
+-cp -p i18n_ref/*.thl i18n; \
++cp -p "$(srcdir)/i18n_ref/"*.thl i18n; \
+   fi
+ 
+ i18n_ref:
+-- 
+2.1.4
+
diff --git a/meta/recipes-extended/texi2html/texi2html_5.0.bb 
b/meta/recipes-extended/texi2html/texi2html_5.0.bb
index eac289e..ae64816 100644
--- a/meta/recipes-extended/texi2html/texi2html_5.0.bb
+++ b/meta/recipes-extended/texi2html/texi2html_5.0.bb
@@ -7,6 +7,7 @@ LIC_FILES_CHKSUM = 
"file://COPYING;md5=59530bdf33659b29e73d4adb9f9f6552"
 PR = "r2"
 
 SRC_URI = "${SAVANNAH_GNU_MIRROR}/texi2html/${BPN}-${PV}.tar.bz2 \
+   file://0001-Allow-compiling-out-of-source.patch \
"
 
 SRC_URI[md5sum] = "f15ac876fcdc8be865b16535f480aa54"
-- 
2.1.4

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

[OE-core] [PATCH] Remove LIC_FILES_CHKSUM from recipes without SRC_URI

2016-10-21 Thread Olaf Mandel 
LICENSE and LIC_FILES_CHKSUM apply to the sources specified by SRC_URI,
not to the recipe itself. As such a license declaration for a source-less
recipe makes little sense. The LICENSE declaration is mandatory, but
LIC_FILES_CHKSUM can be removed in such cases.

Remove the LIC_FILES_CHKSUM declarations from all recipes that do not
need it.

CC: Paul Eggleton <paul.eggle...@linux.intel.com>
Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 meta-selftest/recipes-test/error/error.bb  | 1 -
 meta/recipes-core/meta/buildtools-tarball.bb   | 2 --
 meta/recipes-core/meta/meta-environment.bb | 2 --
 meta/recipes-core/meta/meta-extsdk-toolchain.bb| 2 --
 meta/recipes-core/meta/meta-ide-support.bb | 2 --
 meta/recipes-core/meta/meta-toolchain.bb   | 3 ---
 meta/recipes-core/meta/nativesdk-buildtools-perl-dummy.bb  | 1 -
 meta/recipes-core/meta/signing-keys.bb | 2 --
 meta/recipes-core/meta/testexport-tarball.bb   | 2 --
 meta/recipes-core/meta/uninative-tarball.bb| 2 --
 meta/recipes-core/os-release/os-release.bb | 1 -
 meta/recipes-core/systemd/systemd-compat-units.bb  | 1 -
 meta/recipes-devtools/opkg/opkg-arch-config_1.0.bb | 1 -
 .../postinst-intercept/nativesdk-postinst-intercept_1.0.bb | 1 -
 meta/recipes-devtools/qemu/qemuwrapper-cross_1.0.bb| 1 -
 meta/recipes-devtools/sgml-common/sgml-common-native_0.6.3.bb  | 1 -
 meta/recipes-kernel/kmod/depmodwrapper-cross_1.0.bb| 1 -
 meta/recipes-kernel/linux/kernel-devsrc.bb | 1 -
 meta/recipes-kernel/perf/perf.bb   | 1 -
 19 files changed, 28 deletions(-)

diff --git a/meta-selftest/recipes-test/error/error.bb 
b/meta-selftest/recipes-test/error/error.bb
index a7bdecf..3c22e7c 100644
--- a/meta-selftest/recipes-test/error/error.bb
+++ b/meta-selftest/recipes-test/error/error.bb
@@ -1,7 +1,6 @@
 SUMMARY = "Error Test case that fails on do_compile"
 DESCRIPTION = "This generates a compile time error to be used to for testing."
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 INHIBIT_DEFAULT_DEPS = "1" 

 
 EXCLUDE_FROM_WORLD = "1"
diff --git a/meta/recipes-core/meta/buildtools-tarball.bb 
b/meta/recipes-core/meta/buildtools-tarball.bb
index 5808c95..abdc7fe 100644
--- a/meta/recipes-core/meta/buildtools-tarball.bb
+++ b/meta/recipes-core/meta/buildtools-tarball.bb
@@ -2,8 +2,6 @@ DESCRIPTION = "SDK type target for building a standalone 
tarball containing pyth
tarball can be used to run bitbake builds on systems which 
don't meet the usual version requirements."
 SUMMARY = "Standalone tarball for running builds on systems with inadequate 
software"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 TOOLCHAIN_TARGET_TASK ?= ""
 
diff --git a/meta/recipes-core/meta/meta-environment.bb 
b/meta/recipes-core/meta/meta-environment.bb
index d9e045f..71e0c23 100644
--- a/meta/recipes-core/meta/meta-environment.bb
+++ b/meta/recipes-core/meta/meta-environment.bb
@@ -1,6 +1,4 @@
 SUMMARY = "Package of environment files for SDK"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 LICENSE = "MIT"
 PR = "r8"
 
diff --git a/meta/recipes-core/meta/meta-extsdk-toolchain.bb 
b/meta/recipes-core/meta/meta-extsdk-toolchain.bb
index 2bd9a5b..22de0f3 100644
--- a/meta/recipes-core/meta/meta-extsdk-toolchain.bb
+++ b/meta/recipes-core/meta/meta-extsdk-toolchain.bb
@@ -1,8 +1,6 @@
 SUMMARY = "Extensible SDK toolchain meta-recipe"
 DESCRIPTION = "Meta-recipe for ensuring the build directory contains all 
appropriate toolchain packages for using an IDE"
 LICENSE = "MIT"
-LIC_FILES_CHKSUM = 
"file://${COREBASE}/LICENSE;md5=4d92cd373abda3937c2bc47fbc49d690 \
-
file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420"
 
 DEPENDS = "virtual/libc gdb-cross-${TARGET_ARCH} qemu-native 
qemu-helper-native unfs3-native"
 
diff --git a/meta/recipes-core/meta/meta-ide-support.bb 
b/meta/reci

Re: [OE-core] [PATCH 0/1] $COREBASE/LICENSE in LIC_FILES_CHKSUM

2016-10-21 Thread Olaf Mandel 
Hi Paul,

Am 21.10.2016 um 01:29 schrieb Paul Eggleton:
> On Wed, 19 Oct 2016 10:05:12 Olaf Mandel wrote:
-Snipp-
>> There are additional recipes that reference the global LICENSE
>> file, but where I am not sure if a less blanket license declaration
>> is appropriate or not. Especially for meta-toolchain and
>> build-appliance-image, the global LICENSE is probably correct.
>> Any comments?
> 
> For images and other recipes that don't pull in anything by themselves, there
> shouldn't be a need for LIC_FILES_CHKSUM (or indeed LICENSE) to be set at all
> - in order to understand the license for the result the appropriate place to
> look is the license manifest that represents everything that goes into the
> image. In fact, in master / morty, if SRC_URI is empty then LIC_FILES_CHKSUM
> is no longer verified [1]. Thus, for such recipes we can remove the setting of
> LIC_FILES_CHKSUM (and LICENSE, if possible) altogether.
> 
That seems to apply to eight of the nine remaining recipes still
containing the ${COREBASE}/LICENSE:

meta/recipes-core/meta/buildtools-tarball.bb
meta/recipes-core/meta/meta-environment.bb
meta/recipes-core/meta/meta-extsdk-toolchain.bb
meta/recipes-core/meta/meta-ide-support.bb
meta/recipes-core/meta/meta-toolchain.bb
meta/recipes-core/meta/signing-keys.bb
meta/recipes-core/meta/testexport-tarball.bb
meta/recipes-core/meta/uninative-tarball.bb

Of the already modified recipes, this applies to:
meta/recipes-core/systemd/systemd-compat-units.bb

Also, it is not currently possible to remove LICENSE: still causes
build-errors.

I will send a patch that removes LIC_FILES_CHKSUM from a total of 19
files, but please someone review it: it also removes the CHKSUM from
e.g. kernel-devsrc and perf.

This leaves only one instance of ${COREBASE}/LICENSE:
build-appliance-image. Someone please say if the blanket license text
from "LICENSE" applies or if the narrower license terms from
"COPYING.MIT" suffice.

Best regards,
Olaf



signature.asc
Description: OpenPGP digital signature
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH] toaster: fix if USER_CLASSES is missing buildstats

2016-10-27 Thread Olaf Mandel 
buildstats not being in USER_CLASSES is no reason to fail every
single task in toaster_collect_task_stats: using a non-existing
variable results in a None argument to os.path.join().

Check the variable availability before trying to use it.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 meta/classes/toaster.bbclass | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/meta/classes/toaster.bbclass b/meta/classes/toaster.bbclass
index 4bddf34..638bb39 100644
--- a/meta/classes/toaster.bbclass
+++ b/meta/classes/toaster.bbclass
@@ -157,11 +157,11 @@ python toaster_collect_task_stats() {
 import bb.utils
 import os
 
-toaster_statlist_file = os.path.join(e.data.getVar('BUILDSTATS_BASE', 
True), "toasterstatlist")
-
 if not e.data.getVar('BUILDSTATS_BASE', True):
 return  # if we don't have buildstats, we cannot collect stats
 
+toaster_statlist_file = os.path.join(e.data.getVar('BUILDSTATS_BASE', 
True), "toasterstatlist")
+
 def stat_to_float(value):
 return float(value.strip('% \n\r'))
 
-- 
2.1.4

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

[OE-core] [PATCH] combo-layer: handle file_exclude matching dirs

2016-10-17 Thread Olaf Mandel 
If file_exclude matches a directory, os.unlink() got called with this
directory as an argument.

Filter out paths that end in a directory separator.

This still leaves the (then empty) directories, but this does not affect
the git commit.
---
 scripts/combo-layer | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/scripts/combo-layer b/scripts/combo-layer
index 089b65f..d04d88b 100755
--- a/scripts/combo-layer
+++ b/scripts/combo-layer
@@ -294,6 +294,8 @@ def action_init(conf, args):
 # again. Uses the list of files created by tar (easier
 # than walking the tree).
 for file in files.split('\n'):
+if file.endswith(os.path.sep):
+continue
 for pattern in exclude_patterns:
 if fnmatch.fnmatch(file, pattern):
 os.unlink(os.path.join(*([extract_dir] + 
['..'] * subdir_components + [file])))
-- 
2.1.4

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

[OE-core] [PATCH] (PRE)MIRRORS: add pattern for npm:// without slash

2017-08-24 Thread Olaf Mandel 
For URLs with the npm:// transport but with no other slash in it, the
common MIRRORS and PREMIRRORS pattern of npm://.*/.* fails to match.
Add a new pattern in the mirros.bbclass and own-mirrors.bbclass classes
for these URLs.

Many URLs with the npm:// transport have no slash after the host part:

  npm://registry.npmjs.org;name=foo;version=0.1.2

This means that MIRRORS and PREMIRRORS containing entries like the
first one will not match these URLs:

  npm://.*/.*   # fails to match
  npm://.*  # matches, but not those URLs with a slash

Because the npm:// URLs without a slash seem to be more ubiquitous (they
are generated by recipetool), place the new pattern first.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 meta/classes/mirrors.bbclass | 2 ++
 meta/classes/own-mirrors.bbclass | 1 +
 2 files changed, 3 insertions(+)

diff --git a/meta/classes/mirrors.bbclass b/meta/classes/mirrors.bbclass
index b98684f5c6..710df3c7bf 100644
--- a/meta/classes/mirrors.bbclass
+++ b/meta/classes/mirrors.bbclass
@@ -48,6 +48,7 @@ p4://.*/.*  
http://downloads.yoctoproject.org/mirror/sources/ \n \
 osc://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
 https?$://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
 ftp://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
+npm://.*http://downloads.yoctoproject.org/mirror/sources/ \n \
 npm://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
 cvs://.*/.* http://sources.openembedded.org/ \n \
 svn://.*/.* http://sources.openembedded.org/ \n \
@@ -58,6 +59,7 @@ p4://.*/.*  http://sources.openembedded.org/ \n \
 osc://.*/.* http://sources.openembedded.org/ \n \
 https?$://.*/.* http://sources.openembedded.org/ \n \
 ftp://.*/.* http://sources.openembedded.org/ \n \
+npm://.*http://sources.openembedded.org/ \n \
 npm://.*/.* http://sources.openembedded.org/ \n \
 ${CPAN_MIRROR}  http://cpan.metacpan.org/ \n \
 ${CPAN_MIRROR}  http://search.cpan.org/CPAN/ \n \
diff --git a/meta/classes/own-mirrors.bbclass b/meta/classes/own-mirrors.bbclass
index 12b42675bc..253ed9cb7e 100644
--- a/meta/classes/own-mirrors.bbclass
+++ b/meta/classes/own-mirrors.bbclass
@@ -9,5 +9,6 @@ p4://.*/.*  ${SOURCE_MIRROR_URL}
 osc://.*/.* ${SOURCE_MIRROR_URL}
 https?$://.*/.* ${SOURCE_MIRROR_URL}
 ftp://.*/.* ${SOURCE_MIRROR_URL}
+npm://.*${SOURCE_MIRROR_URL}
 npm://.*/.* ${SOURCE_MIRROR_URL}
 }
-- 
2.11.0

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

Re: [OE-core] [PATCH] (PRE)MIRRORS: add pattern for npm:// without slash

2017-08-25 Thread Olaf Mandel 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

Hello Paul,

Am 24.08.2017 um 22:53 schrieb Paul Eggleton:
> On Friday, 25 August 2017 6:19:46 AM NZST Olaf Mandel wrote:
>> For URLs with the npm:// transport but with no other slash in it,
>> the common MIRRORS and PREMIRRORS pattern of npm://.*/.* fails to
>> match. Add a new pattern in the mirros.bbclass and
>> own-mirrors.bbclass classes for these URLs.
- -Snipp-
> 
> Couldn't we just do away with npm://.*/.* ?
> 
While that would fix my specific problem, it would introduce a
backwards-breaking change: if someone _did_ have a SRC_URI with a
slash in it, this would change where on the mirror the package is
searched for:

SRC_URI = npm://foo.tld/bar;name=baz;version=0.1.2
npm://.*/.* => baz-0.1.2.tar.xz
npm://.*=> bar/baz-0.1.2.tar.xz

So keeping both lines is the better solution. But while testing this I
found that the order I chose (first npm://.*, then npm://.*/.*) is
wrong: this first tries the changed (wrong) location on the mirror and
only then the original (correct) one. It is much more efficient to
sort the lines the other way around: this prevents extra network
traffic, if the mirror is not local. I will send an updated batch soon.

Best regards,
Olaf
- -- 
Olaf Mandel

-BEGIN PGP SIGNATURE-
Version: GnuPG v2

iQEcBAEBCAAGBQJZn+/7AAoJEFS9nbRiUh3/0EMIAJTUoXVsvSLWDUm8phbAEoXb
jQZW7hfP+CHzn1V7X6kTu4CZlQWhxOzYbv07sd2cz/4+Hr2U/7JJCul2QTQGizc/
NxVKaixR8V0tM0irn8NeAHexLXLj+dKdmbvaG/H/BeHhFZc4eQOt6gmbGZ1F7V2g
njcP7nui13HbYpHViGFKMT5hnU9B2gKWQ0FAtY6kGZFS1UDvyGFXyHXn1kU0hzjl
6nRuG0q18uueUEEFxqjo4Cb5BTNEjC/QIGIKbSTsqBIU1r3t7uaPSHIYihnRGZ9M
bCpoFKB2PFgi8qFl+cUaFBmtGMeh+loEIDIi1eZ1HmRO7ST0glDmxB0SKM1N06I=
=QS0a
-END PGP SIGNATURE-
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH v2] (PRE)MIRRORS: add pattern for npm:// without slash

2017-08-25 Thread Olaf Mandel 
For URIs with the npm:// transport but with no other slash in it, the
common MIRRORS and PREMIRRORS pattern of npm://.*/.* fails to match.
Add a new pattern in the mirros.bbclass and own-mirrors.bbclass classes
for these URIs.

Many URIs with the npm:// transport have no slash after the host part:

  npm://registry.npmjs.org;name=foo;version=0.1.2

This means that MIRRORS and PREMIRRORS containing entries like the
first one will not match these URIs:

  npm://.*/.*   # fails to match
  npm://.*  # matches

The npm:// URIs without a slash also matches URIs with a slash, but
then the file is searched for on the mirror with the same path that was
in the URI. Because of that, the pattern with the slash should be first
in the list, to prevent extra requests to the mirrors containing the
wrong path.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 meta/classes/mirrors.bbclass | 2 ++
 meta/classes/own-mirrors.bbclass | 1 +
 2 files changed, 3 insertions(+)

diff --git a/meta/classes/mirrors.bbclass b/meta/classes/mirrors.bbclass
index b98684f5c6..7ac654494b 100644
--- a/meta/classes/mirrors.bbclass
+++ b/meta/classes/mirrors.bbclass
@@ -49,6 +49,7 @@ osc://.*/.* 
http://downloads.yoctoproject.org/mirror/sources/ \n \
 https?$://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
 ftp://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
 npm://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
+npm://.*http://downloads.yoctoproject.org/mirror/sources/ \n \
 cvs://.*/.* http://sources.openembedded.org/ \n \
 svn://.*/.* http://sources.openembedded.org/ \n \
 git://.*/.* http://sources.openembedded.org/ \n \
@@ -59,6 +60,7 @@ osc://.*/.* http://sources.openembedded.org/ \n \
 https?$://.*/.* http://sources.openembedded.org/ \n \
 ftp://.*/.* http://sources.openembedded.org/ \n \
 npm://.*/.* http://sources.openembedded.org/ \n \
+npm://.*http://sources.openembedded.org/ \n \
 ${CPAN_MIRROR}  http://cpan.metacpan.org/ \n \
 ${CPAN_MIRROR}  http://search.cpan.org/CPAN/ \n \
 "
diff --git a/meta/classes/own-mirrors.bbclass b/meta/classes/own-mirrors.bbclass
index 12b42675bc..c70b94e7dc 100644
--- a/meta/classes/own-mirrors.bbclass
+++ b/meta/classes/own-mirrors.bbclass
@@ -10,4 +10,5 @@ osc://.*/.* ${SOURCE_MIRROR_URL}
 https?$://.*/.* ${SOURCE_MIRROR_URL}
 ftp://.*/.* ${SOURCE_MIRROR_URL}
 npm://.*/.* ${SOURCE_MIRROR_URL}
+npm://.*${SOURCE_MIRROR_URL}
 }
-- 
2.11.0

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

Re: [OE-core] [PATCH v2] (PRE)MIRRORS: add pattern for npm:// without slash

2017-08-25 Thread Olaf Mandel 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

Hello Leonardo,

Am 25.08.2017 um 16:32 schrieb Leonardo Sandoval:
> instead of adding a new npm regex, not sure if below one would the
> same job
> 
> npm://.*(/.*)?
> 
No: the string is split into different parts before the RE is
considered. So the fetch step fails with this error:

Exception: sre_constants.error: missing ), unterminated subpattern at
position 2

Probably because the pattern was split to:

['npm', '.*(', '/.*)?', '', '', OrderedDict()]

Best regards,
Olaf
- -- 
Olaf Mandel

-BEGIN PGP SIGNATURE-
Version: GnuPG v2

iQEcBAEBCAAGBQJZoEzpAAoJEFS9nbRiUh3/KgEIAKjWmkwrlR3CXNk1zYgS9XKb
FQsE47IYogdtLMSgjgbjo+tUVj34a4NpPY4kNXdwadXNSOQZJPDcu2hzkvxAyAVd
9pX0YtxASx4A25/kgBKiTzOL62gGf3vYAnCg3hCC4nPnZE5d49Xg/oNvhQ2E34xv
vZMaohiPwK5B2QJpPDYtVXEC72ExoMzeoEy1x46hIplQQTh3CE481TKHbFj1he/a
UVwRalNhePQbpScDu49oYdrg8bIl+GfmuDwFE5zLAWBdkkgFaDlLdbwUoz+om3xa
GUHA9eL0lu8CWmrZe1Y4RZuGgj42yJd1CDd7NA3SkeHcB6gsAPzwyn6/UF2rivo=
=mQeF
-END PGP SIGNATURE-
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH v3] (PRE)MIRRORS: fix pattern for npm:// without slash

2017-08-25 Thread Olaf Mandel 
For URIs with the npm:// transport but with no other slash in it, the
common MIRRORS and PREMIRRORS pattern of npm://.*/.* fails to match.
Make the last slash in the pattern optional in the mirros.bbclass and
own-mirrors.bbclass classes.

Many URIs with the npm:// transport have no slash after the host part:

  npm://registry.npmjs.org;name=foo;version=0.1.2

This means that MIRRORS and PREMIRRORS containing entries like the
first one will not match these URIs:

  npm://.*/.*   # fails to match
  npm://.*/?.*  # matches this and URIs with path components

For normal regular expressions, a pattern like 'npm://.*(/.*)?' would
probably be preferred, but that won't work here: the pattern gets split
into the substrings 'npm', '.*(' and '/.*)?', which are not valid
regular expressions individually.

Signed-off-by: Olaf Mandel <o.man...@menlosystems.com>
---
 meta/classes/mirrors.bbclass | 4 ++--
 meta/classes/own-mirrors.bbclass | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/meta/classes/mirrors.bbclass b/meta/classes/mirrors.bbclass
index b98684f5c6..766f1cb6fa 100644
--- a/meta/classes/mirrors.bbclass
+++ b/meta/classes/mirrors.bbclass
@@ -48,7 +48,7 @@ p4://.*/.*  
http://downloads.yoctoproject.org/mirror/sources/ \n \
 osc://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
 https?$://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
 ftp://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
-npm://.*/.* http://downloads.yoctoproject.org/mirror/sources/ \n \
+npm://.*/?.*http://downloads.yoctoproject.org/mirror/sources/ \n \
 cvs://.*/.* http://sources.openembedded.org/ \n \
 svn://.*/.* http://sources.openembedded.org/ \n \
 git://.*/.* http://sources.openembedded.org/ \n \
@@ -58,7 +58,7 @@ p4://.*/.*  http://sources.openembedded.org/ \n \
 osc://.*/.* http://sources.openembedded.org/ \n \
 https?$://.*/.* http://sources.openembedded.org/ \n \
 ftp://.*/.* http://sources.openembedded.org/ \n \
-npm://.*/.* http://sources.openembedded.org/ \n \
+npm://.*/?.*http://sources.openembedded.org/ \n \
 ${CPAN_MIRROR}  http://cpan.metacpan.org/ \n \
 ${CPAN_MIRROR}  http://search.cpan.org/CPAN/ \n \
 "
diff --git a/meta/classes/own-mirrors.bbclass b/meta/classes/own-mirrors.bbclass
index 12b42675bc..0296d545be 100644
--- a/meta/classes/own-mirrors.bbclass
+++ b/meta/classes/own-mirrors.bbclass
@@ -9,5 +9,5 @@ p4://.*/.*  ${SOURCE_MIRROR_URL}
 osc://.*/.* ${SOURCE_MIRROR_URL}
 https?$://.*/.* ${SOURCE_MIRROR_URL}
 ftp://.*/.* ${SOURCE_MIRROR_URL}
-npm://.*/.* ${SOURCE_MIRROR_URL}
+npm://.*/?.*${SOURCE_MIRROR_URL}
 }
-- 
2.11.0

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

Re: [OE-core] [PATCH v2] (PRE)MIRRORS: add pattern for npm:// without slash

2017-08-25 Thread Olaf Mandel 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

Hello again,

Am 25.08.2017 um 18:14 schrieb Olaf Mandel:
> Am 25.08.2017 um 16:32 schrieb Leonardo Sandoval:
>> instead of adding a new npm regex, not sure if below one would
>> the same job
> 
>> npm://.*(/.*)?
> 
> No: the string is split into different parts before the RE is 
> considered. [...]
> 
- -Snipp-
> ['npm', '.*(', '/.*)?', '', '', OrderedDict()]
> 
So npm://.*(/.*)? does not work. But npm://.*/?.* works and _is_ more
elegant than the two pattern lines from my last patches.

Thank you for the suggestion, I will send a new patch.

Best regards,
Olaf
- -- 
Olaf Mandel
-BEGIN PGP SIGNATURE-
Version: GnuPG v2

iQEcBAEBCAAGBQJZoFehAAoJEFS9nbRiUh3/FzsH/06OUDbQPrf3yrH5OBMJe+wF
5Ky8SUworYqA9uGmzR+3nGqmlxlLlhZFMCsM4+a00nr7kSNLu0KF+BISG9fJT5Up
6PJKDZZBChWYHqhkO8CgB3caiQiJKTlFtKmVznNBzJuIPf+ffTiUb0OX7RgFoHHP
jSuEXqKxQTUJO1k1tGhD53F6+W0rzfk2NXqtI2xfURCwZFcTHvrIYHflwoGZzF3F
gc5J37x5oFSxb3Lfy711ZWDTCOljFzoecdpKbhIj2qPr8sO5WVOTE07/FzW4Mb3P
KUlOpbR3D5eSma8f+lacqNQVx0toWRo15yHOVWssLbcoV6V8rzUu6sF1T1Od7Tw=
=RJXU
-END PGP SIGNATURE-
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH v2] npm: change install directory to upstream default

2018-10-22 Thread Olaf Mandel 
The node binary searches for packages in a number of locations, the last
of which is $PREFIX/lib/node (here: /usr/lib/node) from the list of
GLOBAL_FOLDERS [1]. So change the installation directory for all
packages depending on npm.bbclass to that location. This removes the
need to define the NODE_PATH variable to the non-standard
/usr/lib/node_modules value.

While the Tips for Package Managers [2] discusses installing packages to
/usr/lib/node_modules//, this has several drawbacks:

 * it does not work for the REPL as mentioned in the documentation
 * it also does not work for any code _not_ installed as a global
   package under /usr/lib/node_modules (e.g. /usr/share/foo.js will not
   find any packages below /usr/lib)
 * using the non-default location and then having to set NODE_PATH
   barely saves any time: there are only two file-system lookups (to the
   legacy $HOME/.node_modules and $HOME/.node_libraries) directories
   before the library would be found

And the suggestion was made in the context of deduping the node_modules
tree by installing all packages in a flat hierarchy and using symlinks
to the correct version of each dependency. This is not what OpenEmbedded
does, so none of those benefits (deduping, cleaner packages) are being
had by shifting the installation directory to /usr/lib/node_modules.

The choice of a "proper" installation path is not helped by npm
installing to /usr/lib/node_modules if asked to install globally. Still,
using the location expected by nodejs (/usr/lib/node) seems the right
choice.

[1]: https://nodejs.org/api/modules.html#modules_loading_from_the_global_folders
[2]: https://nodejs.org/api/modules.html#modules_addenda_package_manager_tips

Signed-off-by: Olaf Mandel 
---
 meta/classes/npm.bbclass | 11 ++-
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/meta/classes/npm.bbclass b/meta/classes/npm.bbclass
index c351ff0866..30febcffb2 100644
--- a/meta/classes/npm.bbclass
+++ b/meta/classes/npm.bbclass
@@ -10,7 +10,7 @@ def node_pkgname(d):
 
 NPMPN ?= "${@node_pkgname(d)}"
 
-NPM_INSTALLDIR = "${D}${libdir}/node_modules/${NPMPN}"
+NPM_INSTALLDIR = "${libdir}/node/${NPMPN}"
 
 # function maps arch names to npm arch names
 def npm_oe_arch_map(target_arch, d):
@@ -52,9 +52,10 @@ npm_do_install() {
# changing the home directory to the working directory, the .npmrc will
# be created in this directory
export HOME=${WORKDIR}
-   mkdir -p ${NPM_INSTALLDIR}/
+   mkdir -p ${D}${libdir}/node_modules
npm pack .
npm install --prefix ${D}${prefix} -g --arch=${NPM_ARCH} 
--target_arch=${NPM_ARCH} --production --no-registry ${NPMPN}-${PV}.tgz
+   mv ${D}${libdir}/node_modules ${D}${libdir}/node
if [ -d ${D}${prefix}/etc ] ; then
# This will be empty
rmdir ${D}${prefix}/etc
@@ -62,13 +63,13 @@ npm_do_install() {
 }
 
 python populate_packages_prepend () {
-instdir = d.expand('${D}${libdir}/node_modules/${NPMPN}')
+instdir = d.expand('${D}${NPM_INSTALLDIR}')
 extrapackages = oe.package.npm_split_package_dirs(instdir)
 pkgnames = extrapackages.keys()
 d.prependVar('PACKAGES', '%s ' % ' '.join(pkgnames))
 for pkgname in pkgnames:
 pkgrelpath, pdata = extrapackages[pkgname]
-pkgpath = '${libdir}/node_modules/${NPMPN}/' + pkgrelpath
+pkgpath = '${NPM_INSTALLDIR}/' + pkgrelpath
 # package names can't have underscores but npm packages sometimes use 
them
 oe_pkg_name = pkgname.replace('_', '-')
 expanded_pkgname = d.expand(oe_pkg_name)
@@ -84,7 +85,7 @@ python populate_packages_prepend () {
 }
 
 FILES_${PN} += " \
-${libdir}/node_modules/${NPMPN} \
+${NPM_INSTALLDIR} \
 "
 
 EXPORT_FUNCTIONS do_compile do_install
-- 
2.11.0

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

[OE-core] [PATCH v3] npm: change install directory to upstream default

2018-10-22 Thread Olaf Mandel 
The node binary searches for packages in a number of locations, the last
of which is $PREFIX/lib/node (here: /usr/lib/node) from the list of
GLOBAL_FOLDERS [1]. So change the installation directory for all
packages depending on npm.bbclass to that location. This removes the
need to define the NODE_PATH variable to the non-standard
/usr/lib/node_modules value.

While the Tips for Package Managers [2] discusses installing packages to
/usr/lib/node_modules//, this has several drawbacks:

 * it does not work for the REPL as mentioned in the documentation
 * it also does not work for any code _not_ installed as a global
   package under /usr/lib/node_modules (e.g. /usr/share/foo.js will not
   find any packages below /usr/lib)
 * using the non-default location and then having to set NODE_PATH
   barely saves any time: there are only two file-system lookups (to the
   legacy $HOME/.node_modules and $HOME/.node_libraries) directories
   before the library would be found

And the suggestion was made in the context of deduping the node_modules
tree by installing all packages in a flat hierarchy and using symlinks
to the correct version of each dependency. This is not what OpenEmbedded
does, so none of those benefits (deduping, cleaner packages) are being
had by shifting the installation directory to /usr/lib/node_modules.

The choice of a "proper" installation path is not helped by npm
installing to /usr/lib/node_modules if asked to install globally. Still,
using the location expected by nodejs (/usr/lib/node) seems the right
choice.

[1]: https://nodejs.org/api/modules.html#modules_loading_from_the_global_folders
[2]: https://nodejs.org/api/modules.html#modules_addenda_package_manager_tips

Signed-off-by: Olaf Mandel 
---
 meta/classes/npm.bbclass | 11 ++-
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/meta/classes/npm.bbclass b/meta/classes/npm.bbclass
index c351ff0866..30febcffb2 100644
--- a/meta/classes/npm.bbclass
+++ b/meta/classes/npm.bbclass
@@ -10,7 +10,7 @@ def node_pkgname(d):
 
 NPMPN ?= "${@node_pkgname(d)}"
 
-NPM_INSTALLDIR = "${D}${libdir}/node_modules/${NPMPN}"
+NPM_INSTALLDIR = "${libdir}/node/${NPMPN}"
 
 # function maps arch names to npm arch names
 def npm_oe_arch_map(target_arch, d):
@@ -52,9 +52,10 @@ npm_do_install() {
# changing the home directory to the working directory, the .npmrc will
# be created in this directory
export HOME=${WORKDIR}
-   mkdir -p ${NPM_INSTALLDIR}/
+   mkdir -p ${D}${libdir}/node_modules
npm pack .
npm install --prefix ${D}${prefix} -g --arch=${NPM_ARCH} 
--target_arch=${NPM_ARCH} --production --no-registry ${NPMPN}-${PV}.tgz
+   mv ${D}${libdir}/node_modules ${D}${libdir}/node
if [ -d ${D}${prefix}/etc ] ; then
# This will be empty
rmdir ${D}${prefix}/etc
@@ -62,13 +63,13 @@ npm_do_install() {
 }
 
 python populate_packages_prepend () {
-instdir = d.expand('${D}${libdir}/node_modules/${NPMPN}')
+instdir = d.expand('${D}${NPM_INSTALLDIR}')
 extrapackages = oe.package.npm_split_package_dirs(instdir)
 pkgnames = extrapackages.keys()
 d.prependVar('PACKAGES', '%s ' % ' '.join(pkgnames))
 for pkgname in pkgnames:
 pkgrelpath, pdata = extrapackages[pkgname]
-pkgpath = '${libdir}/node_modules/${NPMPN}/' + pkgrelpath
+pkgpath = '${NPM_INSTALLDIR}/' + pkgrelpath
 # package names can't have underscores but npm packages sometimes use 
them
 oe_pkg_name = pkgname.replace('_', '-')
 expanded_pkgname = d.expand(oe_pkg_name)
@@ -84,7 +85,7 @@ python populate_packages_prepend () {
 }
 
 FILES_${PN} += " \
-${libdir}/node_modules/${NPMPN} \
+${NPM_INSTALLDIR} \
 "
 
 EXPORT_FUNCTIONS do_compile do_install
-- 
2.11.0

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

Re: [OE-core] ✗ patchtest: failure for npm: change install directory to upstream default (rev3)

2018-10-22 Thread Olaf Mandel 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

Hello,

Am 22.10.2018 um 12:33 schrieb Patchwork:
> * Issue Series does not apply on top of target branch
> [test_series_merge_on_head] Suggested fixRebase your series on
> top of targeted branch Targeted branch  master (currently at
> daba6c5a99)
> 
The v2 of the patch does not apply because the master branch already
contains v1 of the patch. I didn't notice this and resubmitted v2 as
v3 after a rebase (to the sumo-branch, not the master branch). So v2
and v3 are identical except for the parent commit hash.

The problem is: v1 does not actually work. What is the procedure to
handle this? Should I make a patch that works on top of the v1 already
in master?

Thanks,
Olaf
- -- 
Olaf Mandel

-BEGIN PGP SIGNATURE-

iQEzBAEBCAAdFiEErGOsr28GGMKRE/TmVL2dtGJSHf8FAlvNqj8ACgkQVL2dtGJS
Hf+zIgf9EFUbmTpQXt1J1rTdK5EQM0o4Jewb7UzGE4XmlSYvi74hem8kjGpYGlXu
46Ec5DR7DgvBzPc2lAw5npVxgC+2PXovVOpne/xnZtwvPM6rKd+AzNNYCm6n52aZ
eGz67KF10+ypSFz76yQ5sxFIc7qivwKmJuzLuB89rFtZRv5aPV0sxZz36f+Duyoi
s0hgdgKCB9IUiXvzVgeCbC3kTFGDuYFGTsBL+X/4YAvgeBoABUJSxtmjScgArk+N
h1Ph9frGqS/jK1v4nU2MW76w/vwhooiRHBxRm/fffVkJiJYXmIqedYWLbsL872q9
QRCJM68mEf4e9dLOepNXbNsfhs2e+w==
=VLfd
-END PGP SIGNATURE-
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH] npm: change install directory to upstream default

2018-10-19 Thread Olaf Mandel 
The node binary searches for packages in a number of locations, the last
of which is $PREFIX/lib/node (here: /usr/lib/node) from the list of
GLOBAL_FOLDERS [1]. Change the installation directory for all packages
depending on npm.bbclass to that location. This removes the need to
define the NODE_PATH variable to the non-standard /usr/lib/node_modules
value.

While the Tips for Package Managers [2] discusses installing packages to
/usr/lib/node_modules//, this has several drawbacks:

 * it does not work for the REPL as mentioned in the documentation
 * it also does not work for any code _not_ installed as a global
   package under /usr/lib/node_modules (e.g. /usr/share/foo.js will not
   find any packages below /usr/lib)
 * using the non-default location and then having to set NODE_PATH
   barely saves any time: there are only two file-system lookups (to the
   legacy $HOME/.node_modules and $HOME/.node_libraries) directories
   before the library would be found

And the suggestion was made in the context of deduping the node_modules
tree by installing all packages in a flat hierarchy and using symlinks
to the correct version of each dependency. This is not what OpenEmbedded
does, so none of those benefits (deduping, cleaner packages) are being
had by shifting the installation directory to /usr/lib/node_modules.

[1]: https://nodejs.org/api/modules.html#modules_loading_from_the_global_folders
[2]: https://nodejs.org/api/modules.html#modules_addenda_package_manager_tips

Signed-off-by: Olaf Mandel 
---
 meta/classes/npm.bbclass | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/meta/classes/npm.bbclass b/meta/classes/npm.bbclass
index c351ff0866..d5ff0c6d57 100644
--- a/meta/classes/npm.bbclass
+++ b/meta/classes/npm.bbclass
@@ -10,7 +10,7 @@ def node_pkgname(d):
 
 NPMPN ?= "${@node_pkgname(d)}"
 
-NPM_INSTALLDIR = "${D}${libdir}/node_modules/${NPMPN}"
+NPM_INSTALLDIR = "${D}${libdir}/node/${NPMPN}"
 
 # function maps arch names to npm arch names
 def npm_oe_arch_map(target_arch, d):
-- 
2.11.0

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

Re: [OE-core] ✗ patchtest: failure for npm: change install directory to upstream default (rev3)

2018-10-23 Thread Olaf Mandel 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

Hello,

Am 23.10.2018 um 00:45 schrieb Richard Purdie:
> On Mon, 2018-10-22 at 12:45 +0200, Olaf Mandel wrote:
>> [...] What is the procedure to handle this? Should I make a
>> patch that works on top of the v1 already in master?
> 
> Normally I'd suggest a new patch against master. In this case I've 
> created one and queued it in next.
> 
Thank you! And sorry about the noise.

Best regards,
Olaf
- -- 
Olaf Mandel

-BEGIN PGP SIGNATURE-

iQEzBAEBCAAdFiEErGOsr28GGMKRE/TmVL2dtGJSHf8FAlvO+B4ACgkQVL2dtGJS
Hf8tXgf/aOMaNVwsdOd+X5d4v81/cw5cquOiGAHscPBCtqyJnfHUjigBQ8OXzqMq
j01RA0G4ML8iIB4HfWBZVtqANIqsPaflRwhG07ykvX2N/VdiMyJ36FBKYmjmsMlF
3Fjr7BF9V2MOEqu2Ddte1S7OD8DAoU+99oUHUhsPSDiubBd1YZhwu9J0KPRXW8wC
uaUww2X7Ung3IVQnGCM6aU8Mvx3KZoDZ92hkhillRHQr9TTJYzXPdI9H7LgKQmI5
FB5JyFwtqMBZbtjv49Rrnkm0apCf1NzZT64RBknKsPL5Afd1wVJ20x8UeZM6+BWR
gJqXgQk+tsremI4tYji9GBn3ST8Udw==
=jWFl
-END PGP SIGNATURE-
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

Re: [OE-core] [PATCH] npm: change install directory to upstream default

2018-10-29 Thread Olaf Mandel 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

Hello Dan,

Am 26.10.2018 um 19:46 schrieb Dan McGregor:
> On Fri, 19 Oct 2018 at 09:29, Olaf Mandel
>  wrote:
>> -NPM_INSTALLDIR = "${D}${libdir}/node_modules/${NPMPN}" 
>> +NPM_INSTALLDIR = "${D}${libdir}/node/${NPMPN}"
> 
> Ï realise this has already gone în, but should this be 
> ${nonarch_libdir}, or does it use /usr/lib64 or /usr/libx32
> correctly?
> 
For me, libdir resolves to /usr/lib, so I didn't notice any problems.
And the "lib"-part is hardcoded in NodeJS (lib/module.js:
path.resolve(prefixDir, 'lib', 'node')).

So if ${nonarch_libdir} is always ${prefix}/lib even if ${libdir}
isn't, then that should be changed.

Best regards,
Olaf Mandel
-BEGIN PGP SIGNATURE-

iQEzBAEBCAAdFiEErGOsr28GGMKRE/TmVL2dtGJSHf8FAlvW43UACgkQVL2dtGJS
Hf8gLAf/aIQ5z219f1Tqtp5Oc1mBr4IN3T/Uzk6MlKKtqfQGS2KBF7CqWLUKNXT2
/eKa/v7H55yTWC6PSlJjx7HMzJK9mhwZ84LwAQ4rpj6ZBn1RYjv4DBdgPxlQN+1N
4kAoWklpgBut0eNtb9hniwf2VhF7rkI8eLVdNcL1SR7JroDahSnjVGozuE/asxK0
yJHHaz3cbe6QooYKeyi4y6lM1wJfP5Wny0rOCMh3rWvsexf7G7itki0MoRU7dNVv
C2zK+/KVjhYbLWSRijSYjGRFpMzKb9AddBxkpIZ4z7vjAlbbbYN1iw3RdPoiPYhP
rwvtSk3Wwe4McsVUjRaEFxwcdCPQQQ==
=1BcQ
-END PGP SIGNATURE-
-- 
___
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

[OE-core] [PATCH] devtool: deploy-target: allow use of alternate ssh

2019-02-21 Thread Olaf Mandel 
Allow the use of an alternate SSH executable for deploy-target and
undeploy-target. The command can be specified with -e / --ssh-exec.

Signed-off-by: Olaf Mandel 
---
 scripts/lib/devtool/deploy.py | 20 
 1 file changed, 16 insertions(+), 4 deletions(-)

diff --git a/scripts/lib/devtool/deploy.py b/scripts/lib/devtool/deploy.py
index 886004b5d0..f345f31b7b 100644
--- a/scripts/lib/devtool/deploy.py
+++ b/scripts/lib/devtool/deploy.py
@@ -211,6 +211,11 @@ def deploy(args, config, basepath, workspace):
 if not args.show_status:
 extraoptions += ' -q'
 
+scp_sshexec = ''
+ssh_sshexec = 'ssh'
+if args.ssh_exec:
+scp_sshexec = "-S %s" % args.ssh_exec
+ssh_sshexec = args.ssh_exec
 scp_port = ''
 ssh_port = ''
 if args.port:
@@ -238,7 +243,7 @@ def deploy(args, config, basepath, workspace):
 for fpath, fsize in filelist:
 f.write('%s %d\n' % (fpath, fsize))
 # Copy them to the target
-ret = subprocess.call("scp %s %s %s/* %s:%s" % (scp_port, 
extraoptions, tmpdir, args.target, os.path.dirname(tmpscript)), shell=True)
+ret = subprocess.call("scp %s %s %s %s/* %s:%s" % (scp_sshexec, 
scp_port, extraoptions, tmpdir, args.target, os.path.dirname(tmpscript)), 
shell=True)
 if ret != 0:
 raise DevtoolError('Failed to copy script to %s - rerun with 
-s to '
 'get a complete error message' % args.target)
@@ -246,7 +251,7 @@ def deploy(args, config, basepath, workspace):
 shutil.rmtree(tmpdir)
 
 # Now run the script
-ret = exec_fakeroot(rd, 'tar cf - . | ssh  %s %s %s \'sh %s %s %s 
%s\'' % (ssh_port, extraoptions, args.target, tmpscript, args.recipename, 
destdir, tmpfilelist), cwd=recipe_outdir, shell=True)
+ret = exec_fakeroot(rd, 'tar cf - . | %s  %s %s %s \'sh %s %s %s %s\'' 
% (ssh_sshexec, ssh_port, extraoptions, args.target, tmpscript, 
args.recipename, destdir, tmpfilelist), cwd=recipe_outdir, shell=True)
 if ret != 0:
 raise DevtoolError('Deploy failed - rerun with -s to get a 
complete '
 'error message')
@@ -276,6 +281,11 @@ def undeploy(args, config, basepath, workspace):
 if not args.show_status:
 extraoptions += ' -q'
 
+scp_sshexec = ''
+ssh_sshexec = 'ssh'
+if args.ssh_exec:
+scp_sshexec = "-S %s" % args.ssh_exec
+ssh_sshexec = args.ssh_exec
 scp_port = ''
 ssh_port = ''
 if args.port:
@@ -292,7 +302,7 @@ def undeploy(args, config, basepath, workspace):
 with open(os.path.join(tmpdir, os.path.basename(tmpscript)), 'w') as f:
 f.write(shellscript)
 # Copy it to the target
-ret = subprocess.call("scp %s %s %s/* %s:%s" % (scp_port, 
extraoptions, tmpdir, args.target, os.path.dirname(tmpscript)), shell=True)
+ret = subprocess.call("scp %s %s %s %s/* %s:%s" % (scp_sshexec, 
scp_port, extraoptions, tmpdir, args.target, os.path.dirname(tmpscript)), 
shell=True)
 if ret != 0:
 raise DevtoolError('Failed to copy script to %s - rerun with -s to 
'
 'get a complete error message' % args.target)
@@ -300,7 +310,7 @@ def undeploy(args, config, basepath, workspace):
 shutil.rmtree(tmpdir)
 
 # Now run the script
-ret = subprocess.call('ssh %s %s %s \'sh %s %s\'' % (ssh_port, 
extraoptions, args.target, tmpscript, args.recipename), shell=True)
+ret = subprocess.call('%s %s %s %s \'sh %s %s\'' % (ssh_sshexec, ssh_port, 
extraoptions, args.target, tmpscript, args.recipename), shell=True)
 if ret != 0:
 raise DevtoolError('Undeploy failed - rerun with -s to get a complete '
'error message')
@@ -324,6 +334,7 @@ def register_commands(subparsers, context):
 parser_deploy.add_argument('-n', '--dry-run', help='List files to be 
deployed only', action='store_true')
 parser_deploy.add_argument('-p', '--no-preserve', help='Do not preserve 
existing files', action='store_true')
 parser_deploy.add_argument('--no-check-space', help='Do not check for 
available space before deploying', action='store_true')
+parser_deploy.add_argument('-e', '--ssh-exec', help='Executable to use in 
place of ssh')
 parser_deploy.add_argument('-P', '--port', help='Specify port to use for 
connection to the target')
 
 strip_opts = parser_deploy.add_mutually_exclusive_group(required=False)
@@ -346,5 +357,6 @@ def register_commands(subparsers, context):
 parser_undeploy.add_argument('-s', '--show-status', help='Show 
progress/status output', action='store_true')
 parser_undeploy.add_argument('-a', '--all', help='Undeploy all recipes 
deployed on the target', action='store_true')
 parser_undeploy.add_argument('-n', '--dry-run',

[OE-core] [PATCH v2] devtool: deploy-target: support "unsafe" symlinks

2019-02-27 Thread Olaf Mandel 
The bosybox version of tar in sumo considers symlink targets that start
with / or with ../ to be unsafe and refuses to unpack them unless the
EXTRACT_UNSAFE_SYMLINKS environment variable is set to 1.

As even many core packages legitimately contain such links (e.g.
coreutils-locale-*, dropbear, eudev, initscripts, kmod, ...), add the
environment variable to the remote script.
---
 scripts/lib/devtool/deploy.py | 5 +
 1 file changed, 5 insertions(+)

diff --git a/scripts/lib/devtool/deploy.py b/scripts/lib/devtool/deploy.py
index f345f31b7b..dc9b34e0ee 100644
--- a/scripts/lib/devtool/deploy.py
+++ b/scripts/lib/devtool/deploy.py
@@ -114,6 +114,11 @@ def _prepare_remote_script(deploy, verbose=False, 
dryrun=False, undeployall=Fals
 lines.append('rm $3')
 lines.append('mkdir -p `dirname $manifest`')
 lines.append('mkdir -p $2')
+# Busybox tar v1.28.x (plus v1.27.2 with the CVE-2011-5325 patch)
+# requires this env variable. The first release without this issue
+# is thud.
+# FIXME Remove once sumo goes out of support
+lines.append('export EXTRACT_UNSAFE_SYMLINKS=1')
 if verbose:
 lines.append('tar xv -C $2 -f - | tee $manifest')
 else:
-- 
2.11.0

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

[OE-core] [PATCH] devtool: deploy-target: support "unsafe" symlinks

2019-02-26 Thread Olaf Mandel 
The bosybox version of tar considers symlink targets that start with /
or with ../ to be unsafe and refuses to unpack them unless the
EXTRACT_UNSAFE_SYMLINKS environment variable is set to 1.

As even many core packages legitimately contain such links (e.g.
coreutils-locale-*, dropbear, eudev, initscripts, kmod, ...), add the
environment variable to the remote script.
---
 scripts/lib/devtool/deploy.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/scripts/lib/devtool/deploy.py b/scripts/lib/devtool/deploy.py
index f345f31b7b..9617388f90 100644
--- a/scripts/lib/devtool/deploy.py
+++ b/scripts/lib/devtool/deploy.py
@@ -114,6 +114,7 @@ def _prepare_remote_script(deploy, verbose=False, 
dryrun=False, undeployall=Fals
 lines.append('rm $3')
 lines.append('mkdir -p `dirname $manifest`')
 lines.append('mkdir -p $2')
+lines.append('export EXTRACT_UNSAFE_SYMLINKS=1')
 if verbose:
 lines.append('tar xv -C $2 -f - | tee $manifest')
 else:
-- 
2.11.0

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