* [Buildroot] [PATCH 1/4] package/libbytesize: add new package
2020-02-17 14:30 [Buildroot] [PATCH 0/4] udisks bump to version 2.8.4 Giulio Benetti
@ 2020-02-17 14:30 ` Giulio Benetti
2020-04-12 14:00 ` Thomas Petazzoni
2020-02-17 14:30 ` [Buildroot] [PATCH 2/4] package/volume_key: " Giulio Benetti
` (2 subsequent siblings)
3 siblings, 1 reply; 15+ messages in thread
From: Giulio Benetti @ 2020-02-17 14:30 UTC (permalink / raw)
To: buildroot
Libbytesize is a tiny library providing a C "class" for working with
arbitrary big sizes in bytes.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
---
DEVELOPERS | 1 +
package/Config.in | 1 +
package/libbytesize/Config.in | 10 ++++++++++
package/libbytesize/libbytesize.hash | 3 +++
package/libbytesize/libbytesize.mk | 15 +++++++++++++++
5 files changed, 30 insertions(+)
create mode 100644 package/libbytesize/Config.in
create mode 100644 package/libbytesize/libbytesize.hash
create mode 100644 package/libbytesize/libbytesize.mk
diff --git a/DEVELOPERS b/DEVELOPERS
index 5991ee505a..6f2dbbb11a 100644
--- a/DEVELOPERS
+++ b/DEVELOPERS
@@ -1009,6 +1009,7 @@ F: package/xapian/
N: Giulio Benetti <giulio.benetti@benettiengineering.com>
F: package/at/
+F: package/libbytesize/
F: package/libnspr/
F: package/libnss/
F: package/minicom/
diff --git a/package/Config.in b/package/Config.in
index 6201c5d2c7..4d76a5791f 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -1782,6 +1782,7 @@ menu "Other"
source "package/libavl/Config.in"
source "package/libb64/Config.in"
source "package/libbsd/Config.in"
+ source "package/libbytesize/Config.in"
source "package/libcap/Config.in"
source "package/libcap-ng/Config.in"
source "package/libcgroup/Config.in"
diff --git a/package/libbytesize/Config.in b/package/libbytesize/Config.in
new file mode 100644
index 0000000000..a3b086ccbc
--- /dev/null
+++ b/package/libbytesize/Config.in
@@ -0,0 +1,10 @@
+config BR2_PACKAGE_LIBBYTESIZE
+ bool "libbytesize"
+ select BR2_PACKAGE_GMP
+ select BR2_PACKAGE_MPFR
+ select BR2_PACKAGE_PCRE2
+ help
+ A tiny library providing a C "class" for working with
+ arbitrary big sizes in bytes.
+
+ https://github.com/storaged-project/libbytesize/
diff --git a/package/libbytesize/libbytesize.hash b/package/libbytesize/libbytesize.hash
new file mode 100644
index 0000000000..2348ae2d25
--- /dev/null
+++ b/package/libbytesize/libbytesize.hash
@@ -0,0 +1,3 @@
+# Locally calculated
+sha256 e2063d5123fca5c347d561a032e406b9b72b94f3c30f7159bb4619c5aef02ef4 libbytesize-2.2.tar.gz
+sha256 97bdc721d875501b6243a456333fdfdb1ab64d31c4da2554de845caf4674b946 LICENSE
diff --git a/package/libbytesize/libbytesize.mk b/package/libbytesize/libbytesize.mk
new file mode 100644
index 0000000000..65cde2a2fc
--- /dev/null
+++ b/package/libbytesize/libbytesize.mk
@@ -0,0 +1,15 @@
+################################################################################
+#
+# libbytesize
+#
+################################################################################
+
+LIBBYTESIZE_VERSION = 2.2
+LIBBYTESIZE_SITE = $(call github,storaged-project,libbytesize,$(LIBBYTESIZE_VERSION))
+LIBBYTESIZE_LICENSE = LGPL-2.1+
+LIBBYTESIZE_LICENSE_FILES = LICENSE
+LIBBYTESIZE_DEPENDENCIES = gmp mpfr pcre2 host-pkgconf
+LIBBYTESIZE_INSTALL_STAGING = YES
+LIBBYTESIZE_AUTORECONF = YES
+
+$(eval $(autotools-package))
--
2.20.1
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Buildroot] [PATCH 1/4] package/libbytesize: add new package
2020-02-17 14:30 ` [Buildroot] [PATCH 1/4] package/libbytesize: add new package Giulio Benetti
@ 2020-04-12 14:00 ` Thomas Petazzoni
0 siblings, 0 replies; 15+ messages in thread
From: Thomas Petazzoni @ 2020-04-12 14:00 UTC (permalink / raw)
To: buildroot
On Mon, 17 Feb 2020 15:30:27 +0100
Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
> Libbytesize is a tiny library providing a C "class" for working with
> arbitrary big sizes in bytes.
>
> Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
> ---
> DEVELOPERS | 1 +
> package/Config.in | 1 +
> package/libbytesize/Config.in | 10 ++++++++++
> package/libbytesize/libbytesize.hash | 3 +++
> package/libbytesize/libbytesize.mk | 15 +++++++++++++++
> 5 files changed, 30 insertions(+)
> create mode 100644 package/libbytesize/Config.in
> create mode 100644 package/libbytesize/libbytesize.hash
> create mode 100644 package/libbytesize/libbytesize.mk
Because the rest of the series has been marked as Changes Requested due
to the feedback that was provided, I have also marked this particular
patch as Changes Requested. It doesn't make a whole lot of sense to
have this libbytesize library without the rest of the series.
Best regards,
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-17 14:30 [Buildroot] [PATCH 0/4] udisks bump to version 2.8.4 Giulio Benetti
2020-02-17 14:30 ` [Buildroot] [PATCH 1/4] package/libbytesize: add new package Giulio Benetti
@ 2020-02-17 14:30 ` Giulio Benetti
2020-02-17 18:51 ` Thomas Petazzoni
2020-02-17 14:30 ` [Buildroot] [PATCH 3/4] package/libblockdev: " Giulio Benetti
2020-02-17 14:30 ` [Buildroot] [PATCH 4/4] package/udisks: bump version to 2.8.4 Giulio Benetti
3 siblings, 1 reply; 15+ messages in thread
From: Giulio Benetti @ 2020-02-17 14:30 UTC (permalink / raw)
To: buildroot
The volume_key project provides a libvolume_key, a library for manipulating
storage volume encryption keys and storing them separately from volumes, and an
associated command-line tool, named volume_key.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
---
DEVELOPERS | 1 +
package/Config.in | 1 +
...ove-po-Makefile.in-entry-from-AC_CON.patch | 28 +++++++++++++++++++
package/volume_key/Config.in | 19 +++++++++++++
package/volume_key/volume_key.hash | 3 ++
package/volume_key/volume_key.mk | 22 +++++++++++++++
6 files changed, 74 insertions(+)
create mode 100644 package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
create mode 100644 package/volume_key/Config.in
create mode 100644 package/volume_key/volume_key.hash
create mode 100644 package/volume_key/volume_key.mk
diff --git a/DEVELOPERS b/DEVELOPERS
index 6f2dbbb11a..5276c09384 100644
--- a/DEVELOPERS
+++ b/DEVELOPERS
@@ -1016,6 +1016,7 @@ F: package/minicom/
F: package/nfs-utils/
F: package/sunxi-mali-mainline/
F: package/sunxi-mali-mainline-driver/
+F: package/volume_key/
N: Gregory Dymarek <gregd72002@gmail.com>
F: package/ding-libs/
diff --git a/package/Config.in b/package/Config.in
index 4d76a5791f..a058d7d2f0 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -1310,6 +1310,7 @@ menu "Crypto"
source "package/tpm2-tss/Config.in"
source "package/trousers/Config.in"
source "package/ustream-ssl/Config.in"
+ source "package/volume_key/Config.in"
source "package/wolfssl/Config.in"
endmenu
diff --git a/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
new file mode 100644
index 0000000000..44d6ee0d4b
--- /dev/null
+++ b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
@@ -0,0 +1,28 @@
+From d1142e667f4f2f2fcd2d36a006919325c1a4377e Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <giulio.benetti@benettiengineering.com>
+Date: Sat, 25 Jan 2020 17:24:38 +0100
+Subject: [PATCH] configure.ac: remove po/Makefile.in entry from
+ AC_CONFIG_FILES
+
+AC_CONFIG_FILES should not contain po/Makefile.in file because
+gettextize will re-add it again causing build failure.
+
+Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
+---
+ configure.ac | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/configure.ac b/configure.ac
+index 7d3d27a..f4295ff 100644
+--- a/configure.ac
++++ b/configure.ac
+@@ -244,5 +244,5 @@ if test "$GCC" = yes ; then
+ fi
+
+ dnl po/Makefile
+-AC_CONFIG_FILES([Makefile contrib/Makefile doc/Makefile lib/Makefile po/Makefile.in python/Makefile python/python/Makefile python/python3/Makefile src/Makefile tests/Makefile])
++AC_CONFIG_FILES([Makefile contrib/Makefile doc/Makefile lib/Makefile python/Makefile python/python/Makefile python/python3/Makefile src/Makefile tests/Makefile])
+ AC_OUTPUT
+--
+2.20.1
+
diff --git a/package/volume_key/Config.in b/package/volume_key/Config.in
new file mode 100644
index 0000000000..0c0827253d
--- /dev/null
+++ b/package/volume_key/Config.in
@@ -0,0 +1,19 @@
+config BR2_PACKAGE_VOLUME_KEY
+ bool "volume_key"
+ depends on BR2_TOOLCHAIN_HAS_THREADS # cryptsetup, libassuan, libglib2, libgpgme, libnss
+ depends on BR2_USE_MMU # cryptsetup, libassuan, libglib2, libgpgme
+ depends on !BR2_STATIC_LIBS # cryptsetup, libnss
+ depends on BR2_TOOLCHAIN_HAS_SYNC_4 # cryptsetup
+ depends on BR2_USE_WCHAR # libglib2
+ depends on BR2_PACKAGE_LIBGPG_ERROR_ARCH_SUPPORTS # libgpgme, libassuan
+ select BR2_PACKAGE_CRYPTSETUP
+ select BR2_PACKAGE_LIBASSUAN
+ select BR2_PACKAGE_LIBGLIB2
+ select BR2_PACKAGE_LIBGPGME
+ select BR2_PACKAGE_LIBNSS
+ help
+ The volume_key project provides a libvolume_key, a library for manipulating
+ storage volume encryption keys and storing them separately from volumes, and an
+ associated command-line tool, named volume_key.
+
+ https://pagure.io/volume_key
diff --git a/package/volume_key/volume_key.hash b/package/volume_key/volume_key.hash
new file mode 100644
index 0000000000..3925429b4e
--- /dev/null
+++ b/package/volume_key/volume_key.hash
@@ -0,0 +1,3 @@
+# Locally calculated
+sha256 f83146352464048f823c164e755dc10da0e06db024de5b1b6f78beeb8438dc22 volume_key-f56614199726ba1897d8687f0927535693b6603f.tar.gz
+sha256 8177f97513213526df2cf6184d8ff986c675afb514d4e68a404010521b880643 COPYING
diff --git a/package/volume_key/volume_key.mk b/package/volume_key/volume_key.mk
new file mode 100644
index 0000000000..8d5bff6d8e
--- /dev/null
+++ b/package/volume_key/volume_key.mk
@@ -0,0 +1,22 @@
+################################################################################
+#
+# volume_key
+#
+################################################################################
+
+VOLUME_KEY_VERSION = f56614199726ba1897d8687f0927535693b6603f
+VOLUME_KEY_SITE = https://pagure.io/volume_key.git
+VOLUME_KEY_SITE_METHOD = git
+VOLUME_KEY_LICENSE = GPL-2.0+
+VOLUME_KEY_LICENSE_FILES = COPYING
+VOLUME_KEY_DEPENDENCIES = cryptsetup libglib2 libgpgme libnss $(TARGET_NLS_DEPENDENCIES)
+VOLUME_KEY_INSTALL_STAGING = YES
+VOLUME_KEY_AUTORECONF = YES
+VOLUME_KEY_GETTEXTIZE = YES
+
+VOLUME_KEY_CONF_OPTS = --with-sysroot=$(STAGING_DIR) \
+ --with-gpgme-prefix=$(STAGING_DIR)/usr \
+ --without-python \
+ --without-python3
+
+$(eval $(autotools-package))
--
2.20.1
^ permalink raw reply related [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-17 14:30 ` [Buildroot] [PATCH 2/4] package/volume_key: " Giulio Benetti
@ 2020-02-17 18:51 ` Thomas Petazzoni
2020-02-18 13:46 ` Giulio Benetti
0 siblings, 1 reply; 15+ messages in thread
From: Thomas Petazzoni @ 2020-02-17 18:51 UTC (permalink / raw)
To: buildroot
Hello,
On Mon, 17 Feb 2020 15:30:28 +0100
Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
> diff --git a/DEVELOPERS b/DEVELOPERS
> index 6f2dbbb11a..5276c09384 100644
> --- a/DEVELOPERS
> +++ b/DEVELOPERS
> @@ -1016,6 +1016,7 @@ F: package/minicom/
> F: package/nfs-utils/
> F: package/sunxi-mali-mainline/
> F: package/sunxi-mali-mainline-driver/
> +F: package/volume_key/
Would you mind naming the package volume-key instead of volume_key ?
> diff --git a/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
> new file mode 100644
> index 0000000000..44d6ee0d4b
> --- /dev/null
> +++ b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
> @@ -0,0 +1,28 @@
> +From d1142e667f4f2f2fcd2d36a006919325c1a4377e Mon Sep 17 00:00:00 2001
> +From: Giulio Benetti <giulio.benetti@benettiengineering.com>
> +Date: Sat, 25 Jan 2020 17:24:38 +0100
> +Subject: [PATCH] configure.ac: remove po/Makefile.in entry from
> + AC_CONFIG_FILES
> +
> +AC_CONFIG_FILES should not contain po/Makefile.in file because
> +gettextize will re-add it again causing build failure.
> +
> +Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Was this reported upstream?
> diff --git a/package/volume_key/Config.in b/package/volume_key/Config.in
> new file mode 100644
> index 0000000000..0c0827253d
> --- /dev/null
> +++ b/package/volume_key/Config.in
> @@ -0,0 +1,19 @@
> +config BR2_PACKAGE_VOLUME_KEY
> + bool "volume_key"
> + depends on BR2_TOOLCHAIN_HAS_THREADS # cryptsetup, libassuan, libglib2, libgpgme, libnss
> + depends on BR2_USE_MMU # cryptsetup, libassuan, libglib2, libgpgme
> + depends on !BR2_STATIC_LIBS # cryptsetup, libnss
> + depends on BR2_TOOLCHAIN_HAS_SYNC_4 # cryptsetup
> + depends on BR2_USE_WCHAR # libglib2
> + depends on BR2_PACKAGE_LIBGPG_ERROR_ARCH_SUPPORTS # libgpgme, libassuan
> + select BR2_PACKAGE_CRYPTSETUP
> + select BR2_PACKAGE_LIBASSUAN
> + select BR2_PACKAGE_LIBGLIB2
> + select BR2_PACKAGE_LIBGPGME
> + select BR2_PACKAGE_LIBNSS
> + help
> + The volume_key project provides a libvolume_key, a library for manipulating
> + storage volume encryption keys and storing them separately from volumes, and an
> + associated command-line tool, named volume_key.
> +
> + https://pagure.io/volume_key
You need a Config.in comment about the (numerous) dependencies.
> diff --git a/package/volume_key/volume_key.mk b/package/volume_key/volume_key.mk
> new file mode 100644
> index 0000000000..8d5bff6d8e
> --- /dev/null
> +++ b/package/volume_key/volume_key.mk
> @@ -0,0 +1,22 @@
> +################################################################################
> +#
> +# volume_key
> +#
> +################################################################################
> +
> +VOLUME_KEY_VERSION = f56614199726ba1897d8687f0927535693b6603f
Why not use the 0.3.12 release ?
> +VOLUME_KEY_SITE = https://pagure.io/volume_key.git
> +VOLUME_KEY_SITE_METHOD = git
> +VOLUME_KEY_LICENSE = GPL-2.0+
I don't see the "or later" option in any of the source files, so shouldn't this be GPL-2.0 ?
> +VOLUME_KEY_LICENSE_FILES = COPYING
> +VOLUME_KEY_DEPENDENCIES = cryptsetup libglib2 libgpgme libnss $(TARGET_NLS_DEPENDENCIES)
> +VOLUME_KEY_INSTALL_STAGING = YES
> +VOLUME_KEY_AUTORECONF = YES
> +VOLUME_KEY_GETTEXTIZE = YES
> +
> +VOLUME_KEY_CONF_OPTS = --with-sysroot=$(STAGING_DIR) \
The --with-sysroot option always looks weird. What is this needed for ?
> + --with-gpgme-prefix=$(STAGING_DIR)/usr \
> + --without-python \
> + --without-python3
> +
> +$(eval $(autotools-package))
In the configure.ac file, I see:
AC_PATH_PROG([GPG], [gpg2])
AC_ARG_VAR([GPG])
AC_DEFINE_UNQUOTED([GPG_PATH], "$GPG", [Path to the gpg2 executable])
So don't we need gnupg on the host for this to work ?
I see:
PKG_CHECK_MODULES(blkid, [blkid])
don't we need the blkid library from util-linux ? I see it is selected
by cryptsetup, which is why it works.
Thanks!
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-17 18:51 ` Thomas Petazzoni
@ 2020-02-18 13:46 ` Giulio Benetti
2020-02-18 14:06 ` Giulio Benetti
2020-02-18 14:46 ` Thomas Petazzoni
0 siblings, 2 replies; 15+ messages in thread
From: Giulio Benetti @ 2020-02-18 13:46 UTC (permalink / raw)
To: buildroot
Hi Thomas,
On 2/17/20 7:51 PM, Thomas Petazzoni wrote:
> Hello,
>
> On Mon, 17 Feb 2020 15:30:28 +0100
> Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
>
>> diff --git a/DEVELOPERS b/DEVELOPERS
>> index 6f2dbbb11a..5276c09384 100644
>> --- a/DEVELOPERS
>> +++ b/DEVELOPERS
>> @@ -1016,6 +1016,7 @@ F: package/minicom/
>> F: package/nfs-utils/
>> F: package/sunxi-mali-mainline/
>> F: package/sunxi-mali-mainline-driver/
>> +F: package/volume_key/
>
> Would you mind naming the package volume-key instead of volume_key ?
Sure
>> diff --git a/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
>> new file mode 100644
>> index 0000000000..44d6ee0d4b
>> --- /dev/null
>> +++ b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
>> @@ -0,0 +1,28 @@
>> +From d1142e667f4f2f2fcd2d36a006919325c1a4377e Mon Sep 17 00:00:00 2001
>> +From: Giulio Benetti <giulio.benetti@benettiengineering.com>
>> +Date: Sat, 25 Jan 2020 17:24:38 +0100
>> +Subject: [PATCH] configure.ac: remove po/Makefile.in entry from
>> + AC_CONFIG_FILES
>> +
>> +AC_CONFIG_FILES should not contain po/Makefile.in file because
>> +gettextize will re-add it again causing build failure.
>> +
>> +Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
>
> Was this reported upstream?
Not this patch, because the issue happens only when gettextizing, and if
instead you build without getextizing it fails. So I don't think
upstream will be interested about it, what do you think?
This was a doubt I had while dealing with this.
>> diff --git a/package/volume_key/Config.in b/package/volume_key/Config.in
>> new file mode 100644
>> index 0000000000..0c0827253d
>> --- /dev/null
>> +++ b/package/volume_key/Config.in
>> @@ -0,0 +1,19 @@
>> +config BR2_PACKAGE_VOLUME_KEY
>> + bool "volume_key"
>> + depends on BR2_TOOLCHAIN_HAS_THREADS # cryptsetup, libassuan, libglib2, libgpgme, libnss
>> + depends on BR2_USE_MMU # cryptsetup, libassuan, libglib2, libgpgme
>> + depends on !BR2_STATIC_LIBS # cryptsetup, libnss
>> + depends on BR2_TOOLCHAIN_HAS_SYNC_4 # cryptsetup
>> + depends on BR2_USE_WCHAR # libglib2
>> + depends on BR2_PACKAGE_LIBGPG_ERROR_ARCH_SUPPORTS # libgpgme, libassuan
>> + select BR2_PACKAGE_CRYPTSETUP
>> + select BR2_PACKAGE_LIBASSUAN
>> + select BR2_PACKAGE_LIBGLIB2
>> + select BR2_PACKAGE_LIBGPGME
>> + select BR2_PACKAGE_LIBNSS
>> + help
>> + The volume_key project provides a libvolume_key, a library for manipulating
>> + storage volume encryption keys and storing them separately from volumes, and an
>> + associated command-line tool, named volume_key.
>> +
>> + https://pagure.io/volume_key
>
> You need a Config.in comment about the (numerous) dependencies.
Oh yes,completely forgotten!
>> diff --git a/package/volume_key/volume_key.mk b/package/volume_key/volume_key.mk
>> new file mode 100644
>> index 0000000000..8d5bff6d8e
>> --- /dev/null
>> +++ b/package/volume_key/volume_key.mk
>> @@ -0,0 +1,22 @@
>> +################################################################################
>> +#
>> +# volume_key
>> +#
>> +################################################################################
>> +
>> +VOLUME_KEY_VERSION = f56614199726ba1897d8687f0927535693b6603f
>
> Why not use the 0.3.12 release ?
Right
>> +VOLUME_KEY_SITE = https://pagure.io/volume_key.git
>> +VOLUME_KEY_SITE_METHOD = git
>> +VOLUME_KEY_LICENSE = GPL-2.0+
>
> I don't see the "or later" option in any of the source files, so shouldn't this be GPL-2.0 ?
Right
>> +VOLUME_KEY_LICENSE_FILES = COPYING
>> +VOLUME_KEY_DEPENDENCIES = cryptsetup libglib2 libgpgme libnss $(TARGET_NLS_DEPENDENCIES)
>> +VOLUME_KEY_INSTALL_STAGING = YES
>> +VOLUME_KEY_AUTORECONF = YES
>> +VOLUME_KEY_GETTEXTIZE = YES
>> +
>> +VOLUME_KEY_CONF_OPTS = --with-sysroot=$(STAGING_DIR) \
>
> The --with-sysroot option always looks weird. What is this needed for ?
This is a mistake, it's been left there after several tries. Sorry.
>> + --with-gpgme-prefix=$(STAGING_DIR)/usr \
>> + --without-python \
>> + --without-python3
>> +
>> +$(eval $(autotools-package))
>
> In the configure.ac file, I see:
>
> AC_PATH_PROG([GPG], [gpg2])
> AC_ARG_VAR([GPG])
> AC_DEFINE_UNQUOTED([GPG_PATH], "$GPG", [Path to the gpg2 executable])
>
> So don't we need gnupg on the host for this to work ?
Yes, host-gnupg2, I've missed it since my distro provides it.
> I see:
>
> PKG_CHECK_MODULES(blkid, [blkid])
>
> don't we need the blkid library from util-linux ? I see it is selected
> by cryptsetup, which is why it works.
Yes, but if cryptsetup already selects it, do I need to repeat it in
volume-key dependencies?
Best regards
--
Giulio Benetti
Benetti Engineering sas
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-18 13:46 ` Giulio Benetti
@ 2020-02-18 14:06 ` Giulio Benetti
2020-02-18 14:47 ` Thomas Petazzoni
2020-02-18 14:46 ` Thomas Petazzoni
1 sibling, 1 reply; 15+ messages in thread
From: Giulio Benetti @ 2020-02-18 14:06 UTC (permalink / raw)
To: buildroot
On 2/18/20 2:46 PM, Giulio Benetti wrote:
> Hi Thomas,
>
> On 2/17/20 7:51 PM, Thomas Petazzoni wrote:
>> Hello,
>>
>> On Mon, 17 Feb 2020 15:30:28 +0100
>> Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
>>
>>> diff --git a/DEVELOPERS b/DEVELOPERS
>>> index 6f2dbbb11a..5276c09384 100644
>>> --- a/DEVELOPERS
>>> +++ b/DEVELOPERS
>>> @@ -1016,6 +1016,7 @@ F: package/minicom/
>>> F: package/nfs-utils/
>>> F: package/sunxi-mali-mainline/
>>> F: package/sunxi-mali-mainline-driver/
>>> +F: package/volume_key/
>>
>> Would you mind naming the package volume-key instead of volume_key ?
>
> Sure
>
>>> diff --git a/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
>>> new file mode 100644
>>> index 0000000000..44d6ee0d4b
>>> --- /dev/null
>>> +++ b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
>>> @@ -0,0 +1,28 @@
>>> +From d1142e667f4f2f2fcd2d36a006919325c1a4377e Mon Sep 17 00:00:00 2001
>>> +From: Giulio Benetti <giulio.benetti@benettiengineering.com>
>>> +Date: Sat, 25 Jan 2020 17:24:38 +0100
>>> +Subject: [PATCH] configure.ac: remove po/Makefile.in entry from
>>> + AC_CONFIG_FILES
>>> +
>>> +AC_CONFIG_FILES should not contain po/Makefile.in file because
>>> +gettextize will re-add it again causing build failure.
>>> +
>>> +Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
>>
>> Was this reported upstream?
>
> Not this patch, because the issue happens only when gettextizing, and if
> instead you build without getextizing it fails. So I don't think
> upstream will be interested about it, what do you think?
> This was a doubt I had while dealing with this.
>
>>> diff --git a/package/volume_key/Config.in b/package/volume_key/Config.in
>>> new file mode 100644
>>> index 0000000000..0c0827253d
>>> --- /dev/null
>>> +++ b/package/volume_key/Config.in
>>> @@ -0,0 +1,19 @@
>>> +config BR2_PACKAGE_VOLUME_KEY
>>> + bool "volume_key"
>>> + depends on BR2_TOOLCHAIN_HAS_THREADS # cryptsetup, libassuan, libglib2, libgpgme, libnss
>>> + depends on BR2_USE_MMU # cryptsetup, libassuan, libglib2, libgpgme
>>> + depends on !BR2_STATIC_LIBS # cryptsetup, libnss
>>> + depends on BR2_TOOLCHAIN_HAS_SYNC_4 # cryptsetup
>>> + depends on BR2_USE_WCHAR # libglib2
>>> + depends on BR2_PACKAGE_LIBGPG_ERROR_ARCH_SUPPORTS # libgpgme, libassuan
>>> + select BR2_PACKAGE_CRYPTSETUP
>>> + select BR2_PACKAGE_LIBASSUAN
>>> + select BR2_PACKAGE_LIBGLIB2
>>> + select BR2_PACKAGE_LIBGPGME
>>> + select BR2_PACKAGE_LIBNSS
>>> + help
>>> + The volume_key project provides a libvolume_key, a library for manipulating
>>> + storage volume encryption keys and storing them separately from volumes, and an
>>> + associated command-line tool, named volume_key.
>>> +
>>> + https://pagure.io/volume_key
>>
>> You need a Config.in comment about the (numerous) dependencies.
>
> Oh yes,completely forgotten!
>
>>> diff --git a/package/volume_key/volume_key.mk b/package/volume_key/volume_key.mk
>>> new file mode 100644
>>> index 0000000000..8d5bff6d8e
>>> --- /dev/null
>>> +++ b/package/volume_key/volume_key.mk
>>> @@ -0,0 +1,22 @@
>>> +################################################################################
>>> +#
>>> +# volume_key
>>> +#
>>> +################################################################################
>>> +
>>> +VOLUME_KEY_VERSION = f56614199726ba1897d8687f0927535693b6603f
>>
>> Why not use the 0.3.12 release ?
>
> Right
>
>>> +VOLUME_KEY_SITE = https://pagure.io/volume_key.git
>>> +VOLUME_KEY_SITE_METHOD = git
>>> +VOLUME_KEY_LICENSE = GPL-2.0+
>>
>> I don't see the "or later" option in any of the source files, so shouldn't this be GPL-2.0 ?
>
> Right
>
>>> +VOLUME_KEY_LICENSE_FILES = COPYING
>>> +VOLUME_KEY_DEPENDENCIES = cryptsetup libglib2 libgpgme libnss $(TARGET_NLS_DEPENDENCIES)
>>> +VOLUME_KEY_INSTALL_STAGING = YES
>>> +VOLUME_KEY_AUTORECONF = YES
>>> +VOLUME_KEY_GETTEXTIZE = YES
>>> +
>>> +VOLUME_KEY_CONF_OPTS = --with-sysroot=$(STAGING_DIR) \
>>
>> The --with-sysroot option always looks weird. What is this needed for ?
>
> This is a mistake, it's been left there after several tries. Sorry.
>
>>> + --with-gpgme-prefix=$(STAGING_DIR)/usr \
>>> + --without-python \
>>> + --without-python3
>>> +
>>> +$(eval $(autotools-package))
>>
>> In the configure.ac file, I see:
>>
>> AC_PATH_PROG([GPG], [gpg2])
>> AC_ARG_VAR([GPG])
>> AC_DEFINE_UNQUOTED([GPG_PATH], "$GPG", [Path to the gpg2 executable])
>>
>> So don't we need gnupg on the host for this to work ?
>
> Yes, host-gnupg2, I've missed it since my distro provides it.
I've retried building, gnupg2 only is supported but it's not mandatory
to build. So as I can understand we can avoid adding host-gnupg2.
>> I see:
>>
>> PKG_CHECK_MODULES(blkid, [blkid])
>>
>> don't we need the blkid library from util-linux ? I see it is selected
>> by cryptsetup, which is why it works.
>
> Yes, but if cryptsetup already selects it, do I need to repeat it in
> volume-key dependencies?
>
> Best regards
>
Best regards
--
Giulio Benetti
Benetti Engineering sas
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-18 14:06 ` Giulio Benetti
@ 2020-02-18 14:47 ` Thomas Petazzoni
2020-02-20 16:28 ` Giulio Benetti
0 siblings, 1 reply; 15+ messages in thread
From: Thomas Petazzoni @ 2020-02-18 14:47 UTC (permalink / raw)
To: buildroot
On Tue, 18 Feb 2020 15:06:47 +0100
Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
> >> AC_PATH_PROG([GPG], [gpg2])
> >> AC_ARG_VAR([GPG])
> >> AC_DEFINE_UNQUOTED([GPG_PATH], "$GPG", [Path to the gpg2 executable])
> >>
> >> So don't we need gnupg on the host for this to work ?
> >
> > Yes, host-gnupg2, I've missed it since my distro provides it.
>
> I've retried building, gnupg2 only is supported but it's not mandatory
> to build. So as I can understand we can avoid adding host-gnupg2.
Weird, I thought AC_PATH_PROG() was aborting if the tool was not found.
If that's not the case, then we should ensure the build never uses
gpg2, even if available, so that the build result doesn't depend on
whether gpg2 is available system-wide or not.
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-18 14:47 ` Thomas Petazzoni
@ 2020-02-20 16:28 ` Giulio Benetti
0 siblings, 0 replies; 15+ messages in thread
From: Giulio Benetti @ 2020-02-20 16:28 UTC (permalink / raw)
To: buildroot
On 2/18/20 3:47 PM, Thomas Petazzoni wrote:
> On Tue, 18 Feb 2020 15:06:47 +0100
> Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
>
>>>> AC_PATH_PROG([GPG], [gpg2])
>>>> AC_ARG_VAR([GPG])
>>>> AC_DEFINE_UNQUOTED([GPG_PATH], "$GPG", [Path to the gpg2 executable])
>>>>
>>>> So don't we need gnupg on the host for this to work ?
>>>
>>> Yes, host-gnupg2, I've missed it since my distro provides it.
>>
>> I've retried building, gnupg2 only is supported but it's not mandatory
>> to build. So as I can understand we can avoid adding host-gnupg2.
>
> Weird, I thought AC_PATH_PROG() was aborting if the tool was not found.
> If that's not the case, then we should ensure the build never uses
> gpg2, even if available, so that the build result doesn't depend on
> whether gpg2 is available system-wide or not.
I went deeper in the volume_key code and found that gnupg2 is needed at
runtime to create packet with passphrase, so even if build doesn't fail
gnupg2 is needed at runtime. In gcr packet is the same.
But then, it turns out that gnupg2's binary name is gpg and not gpg2,
but usually gpg2 name is used(i.e. on Ubuntu 19.10). Also in gcr package
they assume its name is gpg2 but if you try to build you find gpg.
Gnupg2 failure is difficult to find since it's at runtime(i.e. gpgme
calls /usr/bin/gpg2 to encrypt/decrypt something).
So what do you suggest? Better change gpg binary file name to gpg2 or
better fix gcr package to use /usr/bin/gpg instead of /usr/bin/gpg2 And
do the same here with volume-key?
Gcr package is wrong in any case at the moment, so it must be fixed, and
find someone who's able to test it.
Best regards
--
Giulio Benetti
Benetti Engineering sas
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-18 13:46 ` Giulio Benetti
2020-02-18 14:06 ` Giulio Benetti
@ 2020-02-18 14:46 ` Thomas Petazzoni
2020-02-18 14:55 ` Giulio Benetti
1 sibling, 1 reply; 15+ messages in thread
From: Thomas Petazzoni @ 2020-02-18 14:46 UTC (permalink / raw)
To: buildroot
Hello,
On Tue, 18 Feb 2020 14:46:17 +0100
Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
> >> diff --git a/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
> >> new file mode 100644
> >> index 0000000000..44d6ee0d4b
> >> --- /dev/null
> >> +++ b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
> >> @@ -0,0 +1,28 @@
> >> +From d1142e667f4f2f2fcd2d36a006919325c1a4377e Mon Sep 17 00:00:00 2001
> >> +From: Giulio Benetti <giulio.benetti@benettiengineering.com>
> >> +Date: Sat, 25 Jan 2020 17:24:38 +0100
> >> +Subject: [PATCH] configure.ac: remove po/Makefile.in entry from
> >> + AC_CONFIG_FILES
> >> +
> >> +AC_CONFIG_FILES should not contain po/Makefile.in file because
> >> +gettextize will re-add it again causing build failure.
> >> +
> >> +Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
> >
> > Was this reported upstream?
>
> Not this patch, because the issue happens only when gettextizing, and if
> instead you build without getextizing it fails. So I don't think
> upstream will be interested about it, what do you think?
> This was a doubt I had while dealing with this.
Well, it is a problem if one can't autoreconf/gettextize the upstream
project as-is, no?
> > I see:
> >
> > PKG_CHECK_MODULES(blkid, [blkid])
> >
> > don't we need the blkid library from util-linux ? I see it is selected
> > by cryptsetup, which is why it works.
>
> Yes, but if cryptsetup already selects it, do I need to repeat it in
> volume-key dependencies?
I think since the configure.ac explicitly checks it, it makes sense to
not rely on the fact that a dependency by chance also uses it.
Thanks,
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 2/4] package/volume_key: add new package
2020-02-18 14:46 ` Thomas Petazzoni
@ 2020-02-18 14:55 ` Giulio Benetti
0 siblings, 0 replies; 15+ messages in thread
From: Giulio Benetti @ 2020-02-18 14:55 UTC (permalink / raw)
To: buildroot
On 2/18/20 3:46 PM, Thomas Petazzoni wrote:
> Hello,
>
> On Tue, 18 Feb 2020 14:46:17 +0100
> Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
>
>>>> diff --git a/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
>>>> new file mode 100644
>>>> index 0000000000..44d6ee0d4b
>>>> --- /dev/null
>>>> +++ b/package/volume_key/0001-configure.ac-remove-po-Makefile.in-entry-from-AC_CON.patch
>>>> @@ -0,0 +1,28 @@
>>>> +From d1142e667f4f2f2fcd2d36a006919325c1a4377e Mon Sep 17 00:00:00 2001
>>>> +From: Giulio Benetti <giulio.benetti@benettiengineering.com>
>>>> +Date: Sat, 25 Jan 2020 17:24:38 +0100
>>>> +Subject: [PATCH] configure.ac: remove po/Makefile.in entry from
>>>> + AC_CONFIG_FILES
>>>> +
>>>> +AC_CONFIG_FILES should not contain po/Makefile.in file because
>>>> +gettextize will re-add it again causing build failure.
>>>> +
>>>> +Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
>>>
>>> Was this reported upstream?
>>
>> Not this patch, because the issue happens only when gettextizing, and if
>> instead you build without getextizing it fails. So I don't think
>> upstream will be interested about it, what do you think?
>> This was a doubt I had while dealing with this.
>
> Well, it is a problem if one can't autoreconf/gettextize the upstream
> project as-is, no?
Yes, so let's keep it locally.
>>> I see:
>>>
>>> PKG_CHECK_MODULES(blkid, [blkid])
>>>
>>> don't we need the blkid library from util-linux ? I see it is selected
>>> by cryptsetup, which is why it works.
>>
>> Yes, but if cryptsetup already selects it, do I need to repeat it in
>> volume-key dependencies?
>
> I think since the configure.ac explicitly checks it, it makes sense to
> not rely on the fact that a dependency by chance also uses it.
Ok then, it is the first time I add a package with so many dependencies.
Thank you
Best regards
--
Giulio Benetti
Benetti Engineering sas
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 3/4] package/libblockdev: add new package
2020-02-17 14:30 [Buildroot] [PATCH 0/4] udisks bump to version 2.8.4 Giulio Benetti
2020-02-17 14:30 ` [Buildroot] [PATCH 1/4] package/libbytesize: add new package Giulio Benetti
2020-02-17 14:30 ` [Buildroot] [PATCH 2/4] package/volume_key: " Giulio Benetti
@ 2020-02-17 14:30 ` Giulio Benetti
2020-02-17 18:56 ` Thomas Petazzoni
2020-02-17 14:30 ` [Buildroot] [PATCH 4/4] package/udisks: bump version to 2.8.4 Giulio Benetti
3 siblings, 1 reply; 15+ messages in thread
From: Giulio Benetti @ 2020-02-17 14:30 UTC (permalink / raw)
To: buildroot
Libblockdev is a library for manipulating block devices.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
---
DEVELOPERS | 1 +
package/Config.in | 1 +
package/libblockdev/Config.in | 22 +++++++++++++++
package/libblockdev/libblockdev.hash | 3 +++
package/libblockdev/libblockdev.mk | 40 ++++++++++++++++++++++++++++
5 files changed, 67 insertions(+)
create mode 100644 package/libblockdev/Config.in
create mode 100644 package/libblockdev/libblockdev.hash
create mode 100644 package/libblockdev/libblockdev.mk
diff --git a/DEVELOPERS b/DEVELOPERS
index 5276c09384..514b850c08 100644
--- a/DEVELOPERS
+++ b/DEVELOPERS
@@ -1009,6 +1009,7 @@ F: package/xapian/
N: Giulio Benetti <giulio.benetti@benettiengineering.com>
F: package/at/
+F: package/libblockdev/
F: package/libbytesize/
F: package/libnspr/
F: package/libnss/
diff --git a/package/Config.in b/package/Config.in
index a058d7d2f0..40118b58b0 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -1456,6 +1456,7 @@ menu "Hardware handling"
source "package/let-me-create/Config.in"
source "package/libaio/Config.in"
source "package/libatasmart/Config.in"
+ source "package/libblockdev/Config.in"
source "package/libcec/Config.in"
source "package/libfreefare/Config.in"
source "package/libftdi/Config.in"
diff --git a/package/libblockdev/Config.in b/package/libblockdev/Config.in
new file mode 100644
index 0000000000..0b0a943d5f
--- /dev/null
+++ b/package/libblockdev/Config.in
@@ -0,0 +1,22 @@
+config BR2_PACKAGE_LIBBLOCKDEV
+ bool "libblockdev"
+ depends on BR2_PACKAGE_HAS_UDEV
+ depends on BR2_TOOLCHAIN_HAS_THREADS # volume_key -> cryptsetup
+ depends on BR2_USE_MMU # lvm2, volume_key -> cryptsetup
+ depends on !BR2_STATIC_LIBS # volume_key -> cryptsetup
+ depends on BR2_SYSTEM_ENABLE_NLS # volume_key
+ select BR2_PACKAGE_LIBBYTESIZE
+ select BR2_PACKAGE_LIBYAML
+ select BR2_PACKAGE_PARTED
+ select BR2_PACKAGE_VOLUME_KEY
+ select BR2_PACKAGE_UTIL_LINUX_LIBMOUNT
+ select BR2_PACKAGE_UTIL_LINUX
+ help
+ libblockdev is a C library supporting GObject introspection for
+ manipulation of block devices
+
+ https://github.com/storaged-project/libblockdev/
+
+comment "libblockdev needs udev /dev management and a toolchain w/ threads"
+ depends on BR2_USE_MMU
+ depends on !BR2_TOOLCHAIN_HAS_THREADS || !BR2_PACKAGE_HAS_UDEV
diff --git a/package/libblockdev/libblockdev.hash b/package/libblockdev/libblockdev.hash
new file mode 100644
index 0000000000..0bf9a2c934
--- /dev/null
+++ b/package/libblockdev/libblockdev.hash
@@ -0,0 +1,3 @@
+# Locally calculated
+sha256 c5fc030474b44afcaedd8ea10388282534a6d5bb25488b780e645fbf0a6d4d80 libblockdev-2.23-1.tar.gz
+sha256 97bdc721d875501b6243a456333fdfdb1ab64d31c4da2554de845caf4674b946 LICENSE
diff --git a/package/libblockdev/libblockdev.mk b/package/libblockdev/libblockdev.mk
new file mode 100644
index 0000000000..0dcab4c445
--- /dev/null
+++ b/package/libblockdev/libblockdev.mk
@@ -0,0 +1,40 @@
+################################################################################
+#
+# libblockdev
+#
+################################################################################
+
+LIBBLOCKDEV_VERSION = 2.23-1
+LIBBLOCKDEV_SITE = $(call github,storaged-project,libblockdev,$(LIBBLOCKDEV_VERSION))
+LIBBLOCKDEV_LICENSE = LGPL-2.1+
+LIBBLOCKDEV_LICENSE_FILES = LICENSE
+LIBBLOCKDEV_DEPENDENCIES = host-autoconf-archive host-pkgconf kmod \
+ libbytesize libyaml parted udev util-linux \
+ volume_key
+LIBBLOCKDEV_INSTALL_STAGING = YES
+LIBBLOCKDEV_AUTORECONF_OPTS = --include=$(HOST_DIR)/share/autoconf-archive
+LIBBLOCKDEV_AUTORECONF = YES
+
+LIBBLOCKDEV_CONF_OPTS = --without-nvdimm
+
+ifeq ($(BR2_PACKAGE_DMRAID),y)
+LIBBLOCKDEV_CONF_OPTS += --with-dmraid
+LIBBLOCKDEV_DEPENDENCIES += dmraid
+else
+LIBBLOCKDEV_CONF_OPTS += --without-dmraid
+endif
+
+ifeq ($(BR2_PACKAGE_GOBJECT_INTROSPECTION),y)
+LIBBLOCKDEV_CONF_OPTS += --enable-introspection=yes
+LIBBLOCKDEV_DEPENDENCIES += gobject-introspection
+else
+LIBBLOCKDEV_CONF_OPTS += --enable-introspection=no
+endif
+
+# Autoreconf requires an existing m4 directory
+define LIBBLOCKDEV_MKDIR_M4
+ mkdir -p $(@D)/m4
+endef
+LIBBLOCKDEV_POST_PATCH_HOOKS += LIBBLOCKDEV_MKDIR_M4
+
+$(eval $(autotools-package))
--
2.20.1
^ permalink raw reply related [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 3/4] package/libblockdev: add new package
2020-02-17 14:30 ` [Buildroot] [PATCH 3/4] package/libblockdev: " Giulio Benetti
@ 2020-02-17 18:56 ` Thomas Petazzoni
0 siblings, 0 replies; 15+ messages in thread
From: Thomas Petazzoni @ 2020-02-17 18:56 UTC (permalink / raw)
To: buildroot
Hello,
On Mon, 17 Feb 2020 15:30:29 +0100
Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
> diff --git a/package/libblockdev/Config.in b/package/libblockdev/Config.in
> new file mode 100644
> index 0000000000..0b0a943d5f
> --- /dev/null
> +++ b/package/libblockdev/Config.in
> @@ -0,0 +1,22 @@
> +config BR2_PACKAGE_LIBBLOCKDEV
> + bool "libblockdev"
> + depends on BR2_PACKAGE_HAS_UDEV
> + depends on BR2_TOOLCHAIN_HAS_THREADS # volume_key -> cryptsetup
> + depends on BR2_USE_MMU # lvm2, volume_key -> cryptsetup
> + depends on !BR2_STATIC_LIBS # volume_key -> cryptsetup
You're missing a few dependencies from volume-key:
+ depends on BR2_TOOLCHAIN_HAS_SYNC_4 # cryptsetup
+ depends on BR2_PACKAGE_LIBGPG_ERROR_ARCH_SUPPORTS # libgpgme, libassuan
+ depends on BR2_USE_WCHAR # libglib2
Could you double check ?
> + depends on BR2_SYSTEM_ENABLE_NLS # volume_key
There is no such dependency in volume-key.
> + select BR2_PACKAGE_LIBBYTESIZE
> + select BR2_PACKAGE_LIBYAML
> + select BR2_PACKAGE_PARTED
This one depends on BR2_ENABLE_LOCALE
> + select BR2_PACKAGE_VOLUME_KEY
> + select BR2_PACKAGE_UTIL_LINUX_LIBMOUNT
> + select BR2_PACKAGE_UTIL_LINUX
Reverse the order of those last two lines.
> + help
> + libblockdev is a C library supporting GObject introspection for
> + manipulation of block devices
> +
> + https://github.com/storaged-project/libblockdev/
> +
> +comment "libblockdev needs udev /dev management and a toolchain w/ threads"
> + depends on BR2_USE_MMU
> + depends on !BR2_TOOLCHAIN_HAS_THREADS || !BR2_PACKAGE_HAS_UDEV
This is not accurate considering the dependencies of
BR2_PACKAGE_LIBBLOCKDEV.
> +LIBBLOCKDEV_VERSION = 2.23-1
> +LIBBLOCKDEV_SITE = $(call github,storaged-project,libblockdev,$(LIBBLOCKDEV_VERSION))
> +LIBBLOCKDEV_LICENSE = LGPL-2.1+
> +LIBBLOCKDEV_LICENSE_FILES = LICENSE
> +LIBBLOCKDEV_DEPENDENCIES = host-autoconf-archive host-pkgconf kmod \
> + libbytesize libyaml parted udev util-linux \
> + volume_key
kmod is used here, but not selected by the Config.in file. Indentation
is also odd: you should have only one tab to indent the continuation
lines.
> +LIBBLOCKDEV_INSTALL_STAGING = YES
> +LIBBLOCKDEV_AUTORECONF_OPTS = --include=$(HOST_DIR)/share/autoconf-archive
I haven't followed the recent autoreconf rework, is this now the right
way of using autoconf-archive macros ?
> +ifeq ($(BR2_PACKAGE_GOBJECT_INTROSPECTION),y)
> +LIBBLOCKDEV_CONF_OPTS += --enable-introspection=yes
> +LIBBLOCKDEV_DEPENDENCIES += gobject-introspection
> +else
> +LIBBLOCKDEV_CONF_OPTS += --enable-introspection=no
> +endif
We don't have GOI support (yet).
> +# Autoreconf requires an existing m4 directory
> +define LIBBLOCKDEV_MKDIR_M4
> + mkdir -p $(@D)/m4
> +endef
> +LIBBLOCKDEV_POST_PATCH_HOOKS += LIBBLOCKDEV_MKDIR_M4
So the change from Michael Walle that allows to drop this was not
merged yet?
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Buildroot] [PATCH 4/4] package/udisks: bump version to 2.8.4
2020-02-17 14:30 [Buildroot] [PATCH 0/4] udisks bump to version 2.8.4 Giulio Benetti
` (2 preceding siblings ...)
2020-02-17 14:30 ` [Buildroot] [PATCH 3/4] package/libblockdev: " Giulio Benetti
@ 2020-02-17 14:30 ` Giulio Benetti
2020-02-17 19:07 ` Thomas Petazzoni
3 siblings, 1 reply; 15+ messages in thread
From: Giulio Benetti @ 2020-02-17 14:30 UTC (permalink / raw)
To: buildroot
Bump to version 2.8.4 of udisks. All patches have been upstremed, so
drop them all.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
---
DEVELOPERS | 1 +
...-fix-build-with-newer-glibc-versions.patch | 29 -------
.../0002-Fix-systemd-service-file.patch | 33 --------
...issue-with-missing-sys-sysmacros.h-i.patch | 81 -------------------
package/udisks/Config.in | 31 ++++---
package/udisks/udisks.hash | 4 +-
package/udisks/udisks.mk | 23 ++++--
7 files changed, 38 insertions(+), 164 deletions(-)
delete mode 100644 package/udisks/0001-fix-build-with-newer-glibc-versions.patch
delete mode 100644 package/udisks/0002-Fix-systemd-service-file.patch
delete mode 100644 package/udisks/0003-Fix-compilation-issue-with-missing-sys-sysmacros.h-i.patch
diff --git a/DEVELOPERS b/DEVELOPERS
index 514b850c08..6aac009de4 100644
--- a/DEVELOPERS
+++ b/DEVELOPERS
@@ -1018,6 +1018,7 @@ F: package/nfs-utils/
F: package/sunxi-mali-mainline/
F: package/sunxi-mali-mainline-driver/
F: package/volume_key/
+F: package/udisks/
N: Gregory Dymarek <gregd72002@gmail.com>
F: package/ding-libs/
diff --git a/package/udisks/0001-fix-build-with-newer-glibc-versions.patch b/package/udisks/0001-fix-build-with-newer-glibc-versions.patch
deleted file mode 100644
index 336fc440a6..0000000000
--- a/package/udisks/0001-fix-build-with-newer-glibc-versions.patch
+++ /dev/null
@@ -1,29 +0,0 @@
-Fix build with newer glibc versions
-
-This patch has been backported from upstream:
-
-http://cgit.freedesktop.org/udisks/commit/?h=udisks1&id=9829152b12a8924d2e091a00133ed1a3a7ba75c0
-
-Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
-
-From 9829152b12a8924d2e091a00133ed1a3a7ba75c0 Mon Sep 17 00:00:00 2001
-From: Alexandre Rostovtsev <tetromino@gentoo.org>
-Date: Fri, 29 May 2015 21:09:39 -0400
-Subject: fix build with newer glibc versions
-
-https://bugs.freedesktop.org/show_bug.cgi?id=90778
-
-diff --git a/src/helpers/job-drive-detach.c b/src/helpers/job-drive-detach.c
-index eeafcab..d122a1f 100644
---- a/src/helpers/job-drive-detach.c
-+++ b/src/helpers/job-drive-detach.c
-@@ -18,6 +18,7 @@
- *
- */
-
-+#include <sys/stat.h>
- #include <stdio.h>
- #include <string.h>
- #include <errno.h>
---
-cgit v0.10.2
diff --git a/package/udisks/0002-Fix-systemd-service-file.patch b/package/udisks/0002-Fix-systemd-service-file.patch
deleted file mode 100644
index ea347aef73..0000000000
--- a/package/udisks/0002-Fix-systemd-service-file.patch
+++ /dev/null
@@ -1,33 +0,0 @@
-From bf8b4362cef24a6f747e4329305a4939c8f585d1 Mon Sep 17 00:00:00 2001
-From: David King <amigadave@amigadave.com>
-Date: Thu, 2 Jul 2015 13:49:22 +0100
-Subject: [PATCH] Fix systemd service file
-
-udisks-daemon is installed to $(libexecdir), not $(prefix)/lib/udisks.
-
-https://bugzilla.redhat.com/show_bug.cgi?id=1238664
-https://bugs.freedesktop.org/show_bug.cgi?id=91191
-
-[This patch has been backported from upstream:
-https://cgit.freedesktop.org/udisks/commit/?h=udisks1&id=bf8b4362cef24a6f747e4329305a4939c8f585d1]
-Signed-off-by: Pieterjan Camerlynck <pieterjan.camerlynck@gmail.com>
----
- data/Makefile.am | 2 +-
- 1 file changed, 1 insertion(+), 1 deletion(-)
-
-diff --git a/data/Makefile.am b/data/Makefile.am
-index 411ea0f..08af5f4 100644
---- a/data/Makefile.am
-+++ b/data/Makefile.am
-@@ -31,7 +31,7 @@ systemdservicedir = $(systemdsystemunitdir)
- systemdservice_DATA = $(systemdservice_in_files:.service.in=.service)
-
- $(systemdservice_DATA): $(systemdservice_in_files) Makefile
-- @sed -e "s|\@libexecdir\@|$(prefix)/lib/udisks|" $< > $@
-+ @sed -e "s|\@libexecdir\@|$(libexecdir)|" $< > $@
- endif
-
- udevrulesdir = $(slashlibdir)/udev/rules.d
---
-2.7.4
-
diff --git a/package/udisks/0003-Fix-compilation-issue-with-missing-sys-sysmacros.h-i.patch b/package/udisks/0003-Fix-compilation-issue-with-missing-sys-sysmacros.h-i.patch
deleted file mode 100644
index 71152e5de3..0000000000
--- a/package/udisks/0003-Fix-compilation-issue-with-missing-sys-sysmacros.h-i.patch
+++ /dev/null
@@ -1,81 +0,0 @@
-From d934b0fa5ce24a8fd1987b916e435ba991c2ef64 Mon Sep 17 00:00:00 2001
-From: Vadim Kochan <vadim.kochan@petcube.com>
-Date: Mon, 31 Dec 2018 11:00:46 +0200
-Subject: [PATCH] Fix compilation issue due to missing sys/sysmacros.h include
-
-glibc 2.28+ no longer include <sys/sysmacros.h> from <sys/types.h>, so
-<sys/sysmacros.h> must now be explicitly included to use the major,
-minor and makedev macros.
-
-Signed-off-by: Vadim Kochan <vadim.kochan@petcube.com>
----
- src/daemon.c | 1 +
- src/device.c | 1 +
- src/mount-monitor.c | 1 +
- tools/udisks.c | 1 +
- tools/umount-udisks.c | 1 +
- 5 files changed, 5 insertions(+)
-
-diff --git a/src/daemon.c b/src/daemon.c
-index fafcf9a..4303a6d 100644
---- a/src/daemon.c
-+++ b/src/daemon.c
-@@ -45,6 +45,7 @@
- #include <sys/socket.h>
- #include <sys/un.h>
- #include <sys/utsname.h>
-+#include <sys/sysmacros.h>
- #include <net/if_arp.h>
- #include <fcntl.h>
- #include <signal.h>
-diff --git a/src/device.c b/src/device.c
-index 2ae7f38..20920df 100644
---- a/src/device.c
-+++ b/src/device.c
-@@ -33,6 +33,7 @@
- #include <sys/stat.h>
- #include <sys/time.h>
- #include <sys/resource.h>
-+#include <sys/sysmacros.h>
- #include <fcntl.h>
- #include <pwd.h>
- #include <grp.h>
-diff --git a/src/mount-monitor.c b/src/mount-monitor.c
-index 573a69c..0cfa167 100644
---- a/src/mount-monitor.c
-+++ b/src/mount-monitor.c
-@@ -29,6 +29,7 @@
- #include <string.h>
- #include <sys/types.h>
- #include <sys/stat.h>
-+#include <sys/sysmacros.h>
- #include <mntent.h>
-
- #include <glib.h>
-diff --git a/tools/udisks.c b/tools/udisks.c
-index 97e80d7..22a7d87 100644
---- a/tools/udisks.c
-+++ b/tools/udisks.c
-@@ -31,6 +31,7 @@
- #include <sys/types.h>
- #include <sys/stat.h>
- #include <sys/wait.h>
-+#include <sys/sysmacros.h>
- #include <fcntl.h>
- #include <pwd.h>
- #include <grp.h>
-diff --git a/tools/umount-udisks.c b/tools/umount-udisks.c
-index 2813fe0..d915660 100644
---- a/tools/umount-udisks.c
-+++ b/tools/umount-udisks.c
-@@ -30,6 +30,7 @@
- #include <string.h>
- #include <sys/types.h>
- #include <sys/stat.h>
-+#include <sys/sysmacros.h>
- #include <fcntl.h>
- #include <pwd.h>
- #include <grp.h>
---
-2.14.1
-
diff --git a/package/udisks/Config.in b/package/udisks/Config.in
index 04a632dd8d..909b3c8bd4 100644
--- a/package/udisks/Config.in
+++ b/package/udisks/Config.in
@@ -4,21 +4,27 @@ config BR2_PACKAGE_UDISKS
depends on BR2_HOST_GCC_AT_LEAST_4_9 # spidermonkey
depends on BR2_INSTALL_LIBSTDCPP # spidermonkey
depends on BR2_PACKAGE_HAS_UDEV
+ depends on BR2_PACKAGE_HOST_QEMU_ARCH_SUPPORTS # gobject-introspection
depends on BR2_PACKAGE_SPIDERMONKEY_ARCH_SUPPORTS # spidermonkey
- depends on BR2_TOOLCHAIN_GCC_AT_LEAST_4_9 # spidermonkey
+ depends on BR2_SYSTEM_ENABLE_NLS # volume_key
+ depends on BR2_TOOLCHAIN_GCC_AT_LEAST_4_9 # spidermonkey, gobject-introspection
depends on BR2_TOOLCHAIN_HAS_THREADS_NPTL # polkit
- depends on BR2_USE_MMU # lvm2
+ depends on BR2_USE_MMU # lvm2, gobject-introspection
depends on BR2_USE_WCHAR # dbus-glib -> glib2
depends on !BR2_TOOLCHAIN_USES_UCLIBC # polkit, lvm2, parted
depends on !BR2_STATIC_LIBS # lvm2, spidermonkey
+ depends on BR2_TOOLCHAIN_USES_GLIBC
select BR2_PACKAGE_DBUS
select BR2_PACKAGE_DBUS_GLIB
- select BR2_PACKAGE_SG3_UTILS
- select BR2_PACKAGE_POLKIT
- select BR2_PACKAGE_PARTED
- select BR2_PACKAGE_LVM2
+ select BR2_PACKAGE_GOBJECT_INTROSPECTION
select BR2_PACKAGE_LIBATASMART
+ select BR2_PACKAGE_LIBBLOCKDEV
select BR2_PACKAGE_LIBGUDEV
+ select BR2_PACKAGE_LVM2
+ select BR2_PACKAGE_PARTED
+ select BR2_PACKAGE_POLKIT
+ select BR2_PACKAGE_SG3_UTILS
+ select BR2_PACKAGE_LIBXSLT
help
The udisks project provides
@@ -36,24 +42,25 @@ if BR2_PACKAGE_UDISKS
config BR2_PACKAGE_UDISKS_LVM2
bool "lvm2 support"
# The lvm app library can't compile against musl
- depends on BR2_TOOLCHAIN_USES_GLIBC
select BR2_PACKAGE_LVM2_APP_LIBRARY
help
Enable LVM2 support
-comment "lvm2 support needs a glibc toolchain"
- depends on !BR2_TOOLCHAIN_USES_GLIBC
-
endif
comment "udisks needs udev /dev management"
depends on BR2_USE_MMU
depends on !BR2_PACKAGE_HAS_UDEV
-comment "udisks needs a glibc or musl toolchain with locale, C++, wchar, dynamic library, NPTL, gcc >= 4.9"
+comment "udisks support needs NLS enabled"
+ depends on !BR2_SYSTEM_ENABLE_NLS
+
+comment "udisks needs a glibc toolchain with locale, C++, wchar, dynamic library, NPTL, gcc >= 4.9"
depends on BR2_USE_MMU
depends on BR2_PACKAGE_SPIDERMONKEY_ARCH_SUPPORTS
- depends on !BR2_ENABLE_LOCALE || BR2_TOOLCHAIN_USES_UCLIBC || \
+ depends on BR2_PACKAGE_HOST_QEMU_ARCH_SUPPORTS
+ depends on !BR2_ENABLE_LOCALE || \
+ !BR2_TOOLCHAIN_USES_GLIBC || \
!BR2_INSTALL_LIBSTDCPP || \
BR2_STATIC_LIBS || \
!BR2_TOOLCHAIN_HAS_THREADS_NPTL || \
diff --git a/package/udisks/udisks.hash b/package/udisks/udisks.hash
index b40161e323..f8947b0b09 100644
--- a/package/udisks/udisks.hash
+++ b/package/udisks/udisks.hash
@@ -1,3 +1,3 @@
# Locally calculated
-sha256 f2ec82eb0ea7e01dc299b5b29b3c18cdf861236ec43dcff66b3552b4b31c6f71 udisks-1.0.5.tar.gz
-sha256 7d30f41a79a9b453b4972a6d7e3425ff104b39232891bde8ba659f1adb21f771 COPYING
+sha256 d28367bdacfa039719327e68300b6118427874f2dc53ff7109dedb70a7289819 udisks-2.8.4.tar.bz2
+sha256 98046e932dc6c739001e79d7079a3bd958fc55475dcd917d27f884c0c93525cc COPYING
diff --git a/package/udisks/udisks.mk b/package/udisks/udisks.mk
index fe8932a92e..5b37ac3402 100644
--- a/package/udisks/udisks.mk
+++ b/package/udisks/udisks.mk
@@ -4,28 +4,37 @@
#
################################################################################
-UDISKS_VERSION = 1.0.5
-UDISKS_SITE = http://hal.freedesktop.org/releases
+UDISKS_VERSION = 2.8.4
+UDISKS_SOURCE = udisks-$(UDISKS_VERSION).tar.bz2
+UDISKS_SITE = https://github.com/storaged-project/udisks/releases/download/udisks-$(UDISKS_VERSION)
UDISKS_LICENSE = GPL-2.0+
UDISKS_LICENSE_FILES = COPYING
-# For 0002-Fix-systemd-service-file.patch
-UDISKS_AUTORECONF = YES
UDISKS_DEPENDENCIES = \
sg3_utils \
+ host-autoconf-archive \
host-pkgconf \
+ host-gettext \
udev \
dbus \
dbus-glib \
polkit \
parted \
- lvm2 \
libatasmart \
- libgudev
+ libgudev \
+ libblockdev \
+ libxslt
-UDISKS_CONF_OPTS = --disable-remote-access --disable-man-pages
+UDISKS_CONF_OPTS = \
+ --disable-remote-access \
+ --disable-man \
+ --disable-btrfs \
+ --disable-iscsi \
+ --disable-lsm \
+ --disable-zram
ifeq ($(BR2_PACKAGE_UDISKS_LVM2),y)
+UDISKS_DEPENDENCIES += lvm2
UDISKS_CONF_OPTS += --enable-lvm2
endif
--
2.20.1
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Buildroot] [PATCH 4/4] package/udisks: bump version to 2.8.4
2020-02-17 14:30 ` [Buildroot] [PATCH 4/4] package/udisks: bump version to 2.8.4 Giulio Benetti
@ 2020-02-17 19:07 ` Thomas Petazzoni
0 siblings, 0 replies; 15+ messages in thread
From: Thomas Petazzoni @ 2020-02-17 19:07 UTC (permalink / raw)
To: buildroot
On Mon, 17 Feb 2020 15:30:30 +0100
Giulio Benetti <giulio.benetti@benettiengineering.com> wrote:
> Bump to version 2.8.4 of udisks. All patches have been upstremed, so
> drop them all.
Considering the complexity of the bump, it would be nice to have more
details in the commit log.
> diff --git a/package/udisks/Config.in b/package/udisks/Config.in
> index 04a632dd8d..909b3c8bd4 100644
> --- a/package/udisks/Config.in
> +++ b/package/udisks/Config.in
> @@ -4,21 +4,27 @@ config BR2_PACKAGE_UDISKS
> depends on BR2_HOST_GCC_AT_LEAST_4_9 # spidermonkey
> depends on BR2_INSTALL_LIBSTDCPP # spidermonkey
> depends on BR2_PACKAGE_HAS_UDEV
> + depends on BR2_PACKAGE_HOST_QEMU_ARCH_SUPPORTS # gobject-introspection
> depends on BR2_PACKAGE_SPIDERMONKEY_ARCH_SUPPORTS # spidermonkey
> - depends on BR2_TOOLCHAIN_GCC_AT_LEAST_4_9 # spidermonkey
> + depends on BR2_SYSTEM_ENABLE_NLS # volume_key
There is no such dependency in volume_key.
> + depends on BR2_TOOLCHAIN_GCC_AT_LEAST_4_9 # spidermonkey, gobject-introspection
> depends on BR2_TOOLCHAIN_HAS_THREADS_NPTL # polkit
> - depends on BR2_USE_MMU # lvm2
> + depends on BR2_USE_MMU # lvm2, gobject-introspection
> depends on BR2_USE_WCHAR # dbus-glib -> glib2
> depends on !BR2_TOOLCHAIN_USES_UCLIBC # polkit, lvm2, parted
> depends on !BR2_STATIC_LIBS # lvm2, spidermonkey
> + depends on BR2_TOOLCHAIN_USES_GLIBC
Why is this dependency on glibc needed? Coming from GOI ? Also, it
makes the !BR2_TOOLCHAIN_USES_UCLIBC dependency no longer needed.
In addition, you did not propagate all the dependencies of volume_key I
believe (such as SYNC_4). Could you check this ?
> select BR2_PACKAGE_DBUS
> select BR2_PACKAGE_DBUS_GLIB
> - select BR2_PACKAGE_SG3_UTILS
> - select BR2_PACKAGE_POLKIT
> - select BR2_PACKAGE_PARTED
> - select BR2_PACKAGE_LVM2
> + select BR2_PACKAGE_GOBJECT_INTROSPECTION
> select BR2_PACKAGE_LIBATASMART
> + select BR2_PACKAGE_LIBBLOCKDEV
> select BR2_PACKAGE_LIBGUDEV
> + select BR2_PACKAGE_LVM2
> + select BR2_PACKAGE_PARTED
> + select BR2_PACKAGE_POLKIT
> + select BR2_PACKAGE_SG3_UTILS
> + select BR2_PACKAGE_LIBXSLT
Alphabetic ordering would be nice.
> -comment "udisks needs a glibc or musl toolchain with locale, C++, wchar, dynamic library, NPTL, gcc >= 4.9"
> +comment "udisks support needs NLS enabled"
> + depends on !BR2_SYSTEM_ENABLE_NLS
As said above, I don't think that it is true.
> +comment "udisks needs a glibc toolchain with locale, C++, wchar, dynamic library, NPTL, gcc >= 4.9"
> depends on BR2_USE_MMU
> depends on BR2_PACKAGE_SPIDERMONKEY_ARCH_SUPPORTS
> - depends on !BR2_ENABLE_LOCALE || BR2_TOOLCHAIN_USES_UCLIBC || \
> + depends on BR2_PACKAGE_HOST_QEMU_ARCH_SUPPORTS
> + depends on !BR2_ENABLE_LOCALE || \
> + !BR2_TOOLCHAIN_USES_GLIBC || \
> !BR2_INSTALL_LIBSTDCPP || \
> BR2_STATIC_LIBS || \
> !BR2_TOOLCHAIN_HAS_THREADS_NPTL || \
Of course, this will have to be updated.
Overall, it is really crazy. A stupid daemon that manages storage
devices need a Javascript interpreter and GOI.
There was some discussion at
https://github.com/storaged-project/udisks/issues/615 about running
udisks without polkit.
Or maybe we need to package https://wiki.gentoo.org/wiki/Udevil as an
alternative, but it's not maintained since 2015, so not great at all.
See https://github.com/IgnorantGuru/udevil/commits/master.
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
^ permalink raw reply [flat|nested] 15+ messages in thread