Hi,
2012/6/28 Stutz Sven stutzs...@web.de:
Hello to everyone,
I have an evaluation board (at91sam9263-ek Rev. B) from Atmel and I am
trying now to get a customized OS running on this board.
As I read in the manual, that I need for that purpose a special BSP for
this eval. board.
But
Hi,
2012/3/28 Michael Olbrich m.olbr...@pengutronix.de:
On Sun, Mar 25, 2012 at 08:31:35PM +0200, Remy Bohmer wrote:
These hardcoded paths do not work for a multilib environment where
libraries are typically installed in /lib64 for x86-64 instead of /lib
All correct paths can be found
Let install_copy_toolchain itself figure out where to install the
libgcc libraries and its friends.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v5: no changes compared to v4 (just to keep the series together)
rules/gcclibs.make |6 +++---
1 files changed, 3 insertions(+), 3 deletions
These hardcoded paths do not work for a multilib environment where
libraries are typically installed in /lib64 for x86-64 instead of /lib
All correct paths can be found by interrogating the compiler toolchain
for the location of these files.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v5
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v5: No changes compared to v4
rules/zlib.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/zlib.make b/rules/zlib.make
index 1618e1b..2c15ccb 100644
--- a/rules/zlib.make
+++ b/rules/zlib.make
@@ -48,7 +48,8
is being installed.
In another changeset the support for a 2nd C-library installation will be
added. At least after this change (and a few subsequent ptxdist makefile
fixes) the /lib and /usr/lib area will only contain the contents that belong
there in multilib environment.
Signed-off-by: Remy
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v5: No changes compared to v4
rules/procps.make |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/rules/procps.make b/rules/procps.make
index 7ab3786..44d7ea5 100644
--- a/rules/procps.make
+++ b/rules/procps.make
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v5: No changes compared to v4
rules/ncurses.make | 30 +++---
1 files changed, 15 insertions(+), 15 deletions(-)
diff --git a/rules/ncurses.make b/rules/ncurses.make
index 19247bd..bf25512 100644
--- a/rules/ncurses.make
Hi,
2012/3/27 Michael Olbrich m.olbr...@pengutronix.de:
The code currently in master works with OSELAS.Toolchain and CodeSourcery
toolchains. Your code does _not_ work with OSELAS.Toolchain. But if I
understand you corrctly, it works with CodeSourcery toolchains?
Correct, it works with
Hi,
2012/3/27 Michael Olbrich m.olbr...@pengutronix.de:
I want to run some more tests, but this patch and the following patches
look good now. I'll probably merge them right after the release.
Not sure yet how I want to handle the first two patches...
I just build an OSELAS toolchain, and I
Hi Michael,
2012/3/26 Michael Olbrich m.olbr...@pengutronix.de:
On Sun, Mar 25, 2012 at 07:34:47PM +0200, Remy Bohmer wrote:
2012/3/25 Michael Olbrich m.olbr...@pengutronix.de:
On Sat, Mar 24, 2012 at 08:07:02PM +0100, Remy Bohmer wrote:
ifdef PTXCONF_GLIBC_I18N_BIN_LOCALE
- @$(call
Hi,
2012/3/25 Michael Olbrich m.olbr...@pengutronix.de:
On Sat, Mar 24, 2012 at 08:07:03PM +0100, Remy Bohmer wrote:
CROSS_AUTOCONF_SYSROOT_USR := \
- --prefix=/usr --sysconfdir=/etc --localstatedir=/var
+ --prefix=/usr --sysconfdir=/etc --localstatedir=/var
--libdir=/usr
Hi,
2012/3/25 Michael Olbrich m.olbr...@pengutronix.de:
On Sat, Mar 24, 2012 at 08:07:06PM +0100, Remy Bohmer wrote:
Notice that CROSS_AUTOCONF_USR contains --libdir=/usr/lib{,64} setting.
We want this library in /lib{,64} so we override this setting here
explicitly.
Does it break anything
Hi Michael,
2012/3/25 Michael Olbrich m.olbr...@pengutronix.de:
On Sat, Mar 24, 2012 at 08:07:02PM +0100, Remy Bohmer wrote:
ifdef PTXCONF_GLIBC_I18N_BIN_LOCALE
- @$(call install_copy_toolchain_usr, glibc, bin/locale)
+ @$(call install_copy_toolchain_lib, glibc, bin/locale)
endif
Hi,
2012/3/25 Remy Bohmer li...@bohmer.net:
Hi,
2012/3/25 Michael Olbrich m.olbr...@pengutronix.de:
On Sat, Mar 24, 2012 at 08:07:03PM +0100, Remy Bohmer wrote:
CROSS_AUTOCONF_SYSROOT_USR := \
- --prefix=/usr --sysconfdir=/etc --localstatedir=/var
+ --prefix=/usr --sysconfdir
is being installed.
In another changeset the support for a 2nd C-library installation will be
added. At least after this change (and a few subsequent ptxdist makefile
fixes) the /lib and /usr/lib area will only contain the contents that belong
there in multilib environment.
Signed-off-by: Remy
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/readline.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/readline.make b/rules/readline.make
index d5a8ec0..6e6c55a 100644
--- a/rules/readline.make
+++ b/rules/readline.make
@@ -41,7 +41,8
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/zlib.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/zlib.make b/rules/zlib.make
index 1618e1b..2c15ccb 100644
--- a/rules/zlib.make
+++ b/rules/zlib.make
@@ -48,7 +48,8 @@ ZLIB_CONF_ENV
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/ncurses.make | 30 +++---
1 files changed, 15 insertions(+), 15 deletions(-)
diff --git a/rules/ncurses.make b/rules/ncurses.make
index 19247bd..bf25512 100644
--- a/rules/ncurses.make
+++ b/rules/ncurses.make
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/procps.make |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/rules/procps.make b/rules/procps.make
index 7ab3786..44d7ea5 100644
--- a/rules/procps.make
+++ b/rules/procps.make
@@ -51,7 +51,7 @@ PROCPS_MAKEVARS
Hi,
2012/3/23 Michael Olbrich m.olbr...@pengutronix.de:
On Mon, Mar 19, 2012 at 10:14:46PM +0100, Remy Bohmer wrote:
These hardcoded paths do not work for a multilib environment where
libraries are typically installed in /lib64 for x86-64 instead of /lib
All correct paths can be found
Hi,
2012/3/23 Michael Olbrich m.olbr...@pengutronix.de:
Hi,
I've seen the whole series, but I'm rather busy, and this requires a bit
more thinking. So it may take a while before we get everything in.
On Mon, Mar 19, 2012 at 10:14:44PM +0100, Remy Bohmer wrote:
The ARCH_X86_64 architecture
These hardcoded paths do not work for a multilib environment where
libraries are typically installed in /lib64 for x86-64 instead of /lib
All correct paths can be found by interrogating the compiler toolchain
for the location of these files.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
v3: no changes since v2
rules/zlib.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/zlib.make b/rules/zlib.make
index 1618e1b..2c15ccb 100644
--- a/rules/zlib.make
+++ b/rules/zlib.make
is being installed.
In another changeset the support for a 2nd C-library installation will be
added. At least after this change (and a few subsequent ptxdist makefile
fixes) the /lib and /usr/lib area will only contain the contents that belong
there in multilib environment.
Signed-off-by: Remy
Notice that CROSS_AUTOCONF_USR contains --libdir=/usr/lib{,64} setting.
We want this library in /lib{,64} so we override this setting here
explicitly.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
v3: no changes since v2
rules/util-linux-ng.make |1 +
1 files
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
v3: no changes since v2
rules/readline.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/readline.make b/rules/readline.make
index d5a8ec0..6e6c55a 100644
--- a/rules/readline.make
+++ b
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
v3: no changes since v2
rules/ncurses.make | 30 +++---
1 files changed, 15 insertions(+), 15 deletions(-)
diff --git a/rules/ncurses.make b/rules/ncurses.make
index 19247bd..bf25512 100644
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
v3: no changes since v2
rules/procps.make |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/rules/procps.make b/rules/procps.make
index 7ab3786..44d7ea5 100644
--- a/rules/procps.make
+++ b/rules
Hi,
2012/3/20 Michael Olbrich m.olbr...@pengutronix.de:
- local file=$(for dir in ${pkg_pkg_dir}/{,usr/}lib; do
+ local file=$(for dir in ${pkg_pkg_dir}/{,usr/}lib{,64}; do
I don't like the explicit lib64 here. Next we'll get some debian like
multilib toolchain and it's
Hi,
2012/3/20 Juergen Beisert j...@pengutronix.de:
Just delay the evaluation of the expression to the point of time when it is
really required.
diff --git a/rules/flashbench.make b/rules/flashbench.make
index fcb669e..0ef52d1 100644
--- a/rules/flashbench.make
+++ b/rules/flashbench.make
Hi,
2012/3/20 Bernhard Walle bernh...@bwalle.de:
Sorry, I had the old (unpatched) flashbench.make in the rules directory
of the project...
Will send a patch that uses '=' instead of ':=' which is ready to be
applied.
In that case I will drop my patch. Thanks.
Kind regards,
Remy
--
\*
$(PTXCONF_FLASHBENCH_MAX_BUFSIZE))
Acked-by: Remy Bohmer li...@bohmer.net
Kind regards,
Remy
--
ptxdist mailing list
ptxdist@pengutronix.de
Hi,
diff --git a/rules/pre/Rules.make b/rules/pre/Rules.make
index 1d6da52..d6d40e0 100644
--- a/rules/pre/Rules.make
+++ b/rules/pre/Rules.make
@@ -174,6 +174,15 @@ CROSS_ENV_PROGS := \
$(CROSS_ENV_CPP_FOR_BUILD) \
$(CROSS_ENV_LINK_FOR_BUILD)
+# Outputs 'lib' or 'lib64',
Hi,
2012/3/19 Michael Olbrich m.olbr...@pengutronix.de:
+# Outputs 'lib' or 'lib64', derived from the ld-linux.so.2 from the
compiler toolchain
+CROSS_LIB_DIR := $(shell \
+ dl=$$(echo 'int main(void){return 0;}' | \
+ ${CROSS_CC} ${CROSS_CPPFLAGS} ${CROSS_CFLAGS} -x c -o
This error occurs in case PTXCONF_FLASHBENCH_MAX_BUFSIZE is not defined:
$ ptxdist get
expr: syntax error
make: Nothing to be done for `get'.
Change-Id: I1978d62b8781dbd5090b8a47b8fd8b944e6c79dc
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/flashbench.make |8 +++-
1 files
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
rules/ncurses.make | 30 +++---
1 files changed, 15 insertions(+), 15 deletions(-)
diff --git a/rules/ncurses.make b/rules/ncurses.make
index 19247bd..bf25512 100644
--- a/rules/ncurses.make
+++ b
Let install_copy_toolchain itself figure out where to install the
libgcc libraries and its friends.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
rules/gcclibs.make |6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/rules/gcclibs.make b
The ARCH_X86_64 architecture is added as subtype of ARCH_X86 since it
basically is the same architecture, and a mix of both architectures
will be required to allow multilib support in a later commit.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
platforms
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
rules/procps.make |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/rules/procps.make b/rules/procps.make
index 7ab3786..44d7ea5 100644
--- a/rules/procps.make
+++ b/rules/procps.make
@@ -51,7
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
rules/zlib.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/zlib.make b/rules/zlib.make
index 1618e1b..2c15ccb 100644
--- a/rules/zlib.make
+++ b/rules/zlib.make
@@ -48,7 +48,8
is being installed.
In another changeset the support for a 2nd C-library installation will be
added. At least after this change (and a few subsequent ptxdist makefile
fixes) the /lib and /usr/lib area will only contain the contents that belong
there in multilib environment.
Signed-off-by: Remy
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
rules/readline.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/readline.make b/rules/readline.make
index d5a8ec0..6e6c55a 100644
--- a/rules/readline.make
+++ b/rules/readline.make
Notice that CROSS_AUTOCONF_USR contains --libdir=/usr/lib{,64} setting.
We want this library in /lib{,64} so we override this setting here
explicitly.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: no changes since v1
rules/util-linux-ng.make |1 +
1 files changed, 1 insertions(+), 0
* PTXCONF_TARGET_EXTRA_CFLAGS=-march=core2 -m64
* PTXCONF_TARGET_EXTRA_CXXFLAGS=-march=core2 -m64
* PTXCONF_TARGET_EXTRA_LDFLAGS=-march=core2 -m64
From here you can run ptxdist as usual.
Remy Bohmer (9):
[x86_64] Add x86_64 architecture strings.
[x86_64] libgcc need to be install in /lib64 instead of /lib
is being installed.
In another changeset the support for a 2nd C-library installation will be
added. At least after this change (and a few subsequent ptxdist makefile
fixes) the /lib and /usr/lib area will only contain the contents that belong
there in multilib environment.
Signed-off-by: Remy
Let install_copy_toolchain itself figure out where to install the
libgcc libraries and its friends.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/gcclibs.make |6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/rules/gcclibs.make b/rules/gcclibs.make
index
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/readline.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/readline.make b/rules/readline.make
index d5a8ec0..6e6c55a 100644
--- a/rules/readline.make
+++ b/rules/readline.make
@@ -41,7 +41,8
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/procps.make |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/rules/procps.make b/rules/procps.make
index 7ab3786..44d7ea5 100644
--- a/rules/procps.make
+++ b/rules/procps.make
@@ -51,7 +51,7 @@ PROCPS_MAKEVARS
The ARCH_X86_64 architecture is added as subtype of ARCH_X86 since it
basically is the same architecture, and a mix of both architectures
will be required to allow multilib support in a later commit.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
platforms/architecture.in | 15
These hardcoded paths do not work for a multilib environment where
libraries are typically installed in /lib64 for x86-64 instead of /lib
All correct paths can be found by interrogating the compiler toolchain
for the location of these files.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/zlib.make |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/rules/zlib.make b/rules/zlib.make
index 1618e1b..2c15ccb 100644
--- a/rules/zlib.make
+++ b/rules/zlib.make
@@ -48,7 +48,8 @@ ZLIB_CONF_ENV
Notice that CROSS_AUTOCONF_USR contains --libdir=/usr/lib{,64} setting.
We want this library in /lib{,64} so we override this setting here
explicitly.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/util-linux-ng.make |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/ncurses.make | 30 +++---
1 files changed, 15 insertions(+), 15 deletions(-)
diff --git a/rules/ncurses.make b/rules/ncurses.make
index 19247bd..bf25512 100644
--- a/rules/ncurses.make
+++ b/rules/ncurses.make
Hi,
2012/3/15 Michael Olbrich m.olbr...@pengutronix.de:
On Tue, Mar 13, 2012 at 10:32:14PM +0100, Remy Bohmer wrote:
The platformconfig can specify some valuable global flags for each
package in the configuration to build, for example:
* CPU optimisations (-march= and friends)
* multilib
Hi,
2012/3/14 Michael Olbrich m.olbr...@pengutronix.de:
Indent with on tab everywhere.
OK
+
+BUSYBOX_MAKE_ENV := \
+ $(CROSS_ENV) \
+ CFLAGS=$(CROSS_CFLAGS) -I$(KERNEL_HEADERS_INCLUDE_DIR)
+ SKIP_STRIP=y
Looks like SKIP_STRIP=y must be in BUSYBOX_MAKE_OPT, or
From: Remy Bohmer remy.boh...@oce.com
Busybox make is not properly called in case:
* compiler name tuple does not match target tuple (as used in multilib
compilers, for example: i686-pc-linux-gnu-gcc can compile for
x86_64-unknown-linux-gnu targets. To solve this the 'SUBARCH' need to be
set
to select the proper sysroot to
compile against in busybox
* LDFLAGS to link against the proper sysroot.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: Reworked review comments from Michael Olbrich
v3: Convert spaces-tabs + Added '\' after CFLAGS line
checked after build if build-target
Hi,
2012/3/13 Michael Olbrich m.olbr...@pengutronix.de:
+-CFLAGS = -D_GNU_SOURCE -Wall -Wno-nonnull -Isrc/include
++CFLAGS = -D_GNU_SOURCE -Wall -Wno-nonnull -Isrc/include $(CXXFLAGS)
This should be:
CFLAGS = $(CPPFLAGS) -D_GNU_SOURCE -Wall -Wno-nonnull -Isrc/include $(CFLAGS)
CFLAGS is
Hi,
2012/3/13 Michael Olbrich m.olbr...@pengutronix.de:
lib=${1}
+ extra_cflags=$(source ${PTXDIST_PLATFORMCONFIG} echo
${PTXCONF_TARGET_EXTRA_CFLAGS})
I think ptxd_get_ptxconf should work here. And I think you should use
PTXCONF_TARGET_EXTRA_CFLAGS and
Hi,
2012/3/13 Michael Olbrich m.olbr...@pengutronix.de:
On Mon, Mar 12, 2012 at 10:25:02PM +0100, Remy Bohmer wrote:
BUSYBOX_MAKE_OPT := \
ARCH=$(PTXCONF_ARCH_STRING) \
+ SUBARCH=$(PTXCONF_ARCH_STRING) \
CROSS_COMPILE=$(COMPILER_PREFIX) \
CONFIG_EXTRA_CFLAGS
Hi,
2012/3/13 Michael Olbrich m.olbr...@pengutronix.de:
On Mon, Mar 12, 2012 at 10:25:03PM +0100, Remy Bohmer wrote:
It is assumed by the makefiles that all dynamic linkers must be put in
/lib. This is however not true. Especially on multilib configurations this
information cannot be guessed
Hi,
FYI: With these 5 patches (and a few unfinished architecture.in
patches I have here locally) I at least get the glibc and the dynamic
linkers installed in the right place and all binaries we used to be
compiled with the right flags.
Furthermore, I noticed that the extra flags are not passed
Hi,
2012/3/13 Michael Olbrich m.olbr...@pengutronix.de:
On Tue, Mar 13, 2012 at 01:46:26PM +0100, Remy Bohmer wrote:
It seems I have to stick with the bare minimal options I can set
here... Or any other ideas?
Don't set any flags in BUSYBOX_MAKE_OPT. From what I can tell, the current
need to be passed to GCC here.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: Reworked review comments from Michael Olbrich
scripts/install_copy_toolchain.sh | 10 +++---
1 files changed, 7 insertions(+), 3 deletions(-)
diff --git a/scripts/install_copy_toolchain.sh
b/scripts
there, so remove the unneeded options.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2: repost to keep the series together
rules/glibc.make|2 +-
rules/post/install.make |9 +++--
rules/uclibc.make |2 +-
3 files changed, 5 insertions(+), 8 deletions(-)
diff --git
This link is always being generated and it is assumed that it must
always point to /lib/libdl.so.2.
But this script is not needed and points to the wrong location when a
multilib toolchain is being used where the dynamic linker must be in
/lib64
Signed-off-by: Remy Bohmer li...@bohmer.net
---
v2
:
https://sourcery.mentor.com/sgpp/lite/ia32/portal/doc12648/getting-started.pdf
Signed-off-by: Remy Bohmer li...@bohmer.net
---
patches/rt-tests-0.73/add-extra-cflags.patch | 23 +++
patches/rt-tests-0.73/series |1 +
rules/rt-tests.make
need to be passed to GCC here.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
scripts/install_copy_toolchain.sh |8 +---
1 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/scripts/install_copy_toolchain.sh
b/scripts/install_copy_toolchain.sh
index 8a67698..48d65d7 100755
to select the proper sysroot to
compile against in busybox
* LDFLAGS to link against the proper sysroot.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/busybox.make |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/rules/busybox.make b/rules/busybox.make
index
there, so remove the unneeded options.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/glibc.make|2 +-
rules/post/install.make |9 +++--
rules/uclibc.make |2 +-
3 files changed, 5 insertions(+), 8 deletions(-)
diff --git a/rules/glibc.make b/rules/glibc.make
index
This link is always being generated and it is assumed that it must
always point to /lib/libdl.so.2.
But this script is not needed and points to the wrong location when a
multilib toolchain is being used where the dynamic linker must be in
/lib64
Signed-off-by: Remy Bohmer li...@bohmer.net
on ARCH_ARM
I do not see a reason why you want to change that, but anyway:
Acked by: Remy Bohmer li...@bohmer.net
Kind regards,
Remy
--
ptxdist mailing list
ptxdist@pengutronix.de
2011/3/4 Sascha Hauer s.ha...@pengutronix.de:
They are now unused and should never be introduced again. If anything,
we want to depend on the ARM architecture, but not on specific SOCs.
OK, now I see why you want to do that :-)
Remy
--
ptxdist mailing list
ptxdist@pengutronix.de
Hi,
2010/12/4 Stéphane List sl...@lilotux.net:
Hi,
Is it possible to use the toolchain provided by Ubuntu instead of the
toolchain provided by Oselas ?
I would like to build a small distro for arm with ptxdist.
If you do not want to compile a toolchain yourself, why not use the
toolchains
Hi,
I just pulled the latest git tree of ptxdit with the Kconfig update to
2.6.37-rc2 version.
Now I see the warnings below. I do not think they harm, any idea how
to get rid of them?
---
-- Running: 'ptxdist allyesconfig collection'
Hi,
2010/11/15 Enrico Weigelt weig...@metux.de:
* Remy Bohmer li...@bohmer.net schrieb:
Hi,
big_snip /
I'm now using git as the _primary_ sourcecode management for
about 2 years, and I don't have single patches anymore.
Cool!
My buildsystem (Briegel) directly checks out the source tree
and patch works as usual.
* If a source package does not have a .gitignore file on its own, a default
file is being generated for that package.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
bin/ptxdist |9 +-
config/setup/Kconfig| 42 ++
rules
Hi,
2010/11/1 Michael Olbrich m.olbr...@pengutronix.de:
Hi,
does this happen with the latest git? If there is nothing to
configure/build it shouldn't get that far. Can you send the makefile?
Yes, it happens with latest git.
It happens with source code build out-of-tree according to the
Hi,
I did it with md5 sums of the patches, and even on a heavily patched
package (kernel with 7MB of patches) it can be determined if patches
have changed within about 1 second. Most packages do not have such a
big patchset, so checking will go faster.
For short operations ptxdist is
Hi,
2010/11/1 Michael Olbrich m.olbr...@pengutronix.de:
Hi,
On Mon, Nov 01, 2010 at 02:08:51PM +0100, Remy Bohmer wrote:
2010/11/1 Michael Olbrich m.olbr...@pengutronix.de:
does this happen with the latest git? If there is nothing to
configure/build it shouldn't get that far. Can you send
Hi,
2010/11/1 Michael Olbrich m.olbr...@pengutronix.de:
On Mon, Nov 01, 2010 at 03:44:27PM +0100, Remy Bohmer wrote:
$(STATEDIR)/kernel.install
select KERNEL in io_can.in is possible. There is a dummy symbol for this
in ptxconfig.
Great, this was one issue
Hi,
2010/10/29 Michael Olbrich m.olbr...@pengutronix.de:
Hi,
On Fri, Oct 29, 2010 at 05:03:58PM +0200, Remy Bohmer wrote:
I am figuring out how to make the patching of the packages more
flexible. Currently we can use quilt and git to extract the packages.
I want to:
* Add a dependency
In case the extract stage is empty and PKG_CONF_TOOL is set to NO, then
the prepare stage fails. First a cd was done to the build_dir, but that
does not exist when the extract is empty. This makes the prepare stage fail
while we told it with PKG_CONF_TOOL=NO to skip that stage
Signed-off-by: Remy
Hi,
2010/7/16 Michael Olbrich m.olbr...@pengutronix.de:
On Thu, Jul 15, 2010 at 03:37:52PM +0200, Remy Bohmer wrote:
2010/7/14 Michael Olbrich m.olbr...@pengutronix.de:
On Tue, Jul 13, 2010 at 10:54:09PM +0200, Remy Bohmer wrote:
I just pulled in the latest git and noticed
Hi,
2010/7/16 Michael Olbrich m.olbr...@pengutronix.de:
On Fri, Jul 16, 2010 at 12:21:25PM +0200, Remy Bohmer wrote:
2010/7/16 Michael Olbrich m.olbr...@pengutronix.de:
On Thu, Jul 15, 2010 at 03:37:52PM +0200, Remy Bohmer wrote:
2010/7/14 Michael Olbrich m.olbr...@pengutronix.de:
On Tue
Hi,
2010/7/14 Michael Olbrich m.olbr...@pengutronix.de:
Hi,
On Tue, Jul 13, 2010 at 10:54:09PM +0200, Remy Bohmer wrote:
I just pulled in the latest git and noticed that the '
selected_toolchain' link becomes broken during the build.
Before I start ' ptxdist go' I have all ' selected_
Hi,
I just pulled in the latest git and noticed that the '
selected_toolchain' link becomes broken during the build.
Before I start ' ptxdist go' I have all ' selected_*' links in place
to valid locations.
Then I start 'ptxdist go', which result in this error:
Hi,
What is the intended workflow for the PTXCONF_SETUP_PATCHIN_GIT
developer setting?
I am figuring out how it is supposed to work and if it could (finally)
be used for a workflow like this:
1) Extract source package
2) patch sources while building up git-tree
3) build the tree...
Now here
Hi,
2010/7/1 Juergen Beisert j...@pengutronix.de:
Since today ptxdist-2010.07.0 is available at
http://www.pengutronix.de/software/ptxdist/download/
Comments are welcome, also ideas about topics you miss.
Maybe it could be an idea to make a git tree from the documentation so
that users can:
Hi,
kernel.targetinstall.post. This results in the rebuilds that you can see.
Try this:
deprule=${ptx_state_dir}/${pkg_stamp}: \$(wildcard ${src})
This should give you
kernel.targetinstall.post: file
for kernel-modules.deps and no circular rebuilds.
Ping.
Did you find the time to test
To check for the existence of certain files and include them if these
exist can be done by just simply prepending the include directive with
a '-' sign, without if and wildcard checks.
Signed-off-by: Remy Bohmer li...@bohmer.net
---
rules/other/Toplevel.make | 13 +++--
1 files
are included in the top-level ptxdist makefile
and this makes ptxdist properly handle the dependencies.
Signed-off-by: Remy Bohmer li...@bohmer.net
Signed-off-by: Bart vdr. Meulen bartvdrmeu...@gmail.com
---
rules/other/Toplevel.make|3 +++
scripts/lib/ptxd_make_install.sh |3
The result is an almost empty collection config file with no options selected,
like this:
===
PTXCONF__collectionconfig_MAGIC__=y
PTXCONF_COLLECTION_ALL=y
===
Signed-off-by: Remy Bohmer li...@bohmer.net
---
bin/ptxdist
Hi Michael,
2010/6/20 Michael Olbrich m.olbr...@pengutronix.de:
On Sat, Jun 19, 2010 at 03:54:06PM +0200, Remy Bohmer wrote:
Hi,
2010/6/19 Michael Olbrich m.olbr...@pengutronix.de:
Hi,
Inspired by Remy Bohmer dependencies on installed files, I did the same for
patches and series files
Hi,
2010/6/19 Michael Olbrich m.olbr...@pengutronix.de:
On Sat, Jun 05, 2010 at 12:08:34PM +0200, Remy Bohmer wrote:
When files in the projectroot directory are being used in install_copy() and
install_alternative() then there are no dependencies to these files.
This is being solved
Hi,
2010/6/19 Michael Olbrich m.olbr...@pengutronix.de:
Hi,
Inspired by Remy Bohmer dependencies on installed files, I did the same for
patches and series files.
Adding more dependencies, especially to early build stages is always
tricky. I'd like to avoid unnecessary rebuilds. So please
Hi,
I noticed at the git log these patches:
* [quota-tools] fix menu entry
* [xfsprogs] fix menu indentation
* [lshw] fix menu indentation
* [lvm2] fix menu indention
These patches all seem to do this:
- prompt lvm2
+ prompt lvm2
True, those extra spaces in
Hi,
I followed the same structure as what was now the case of the source
package and makefiles.
Of course it can be done differently, but this worked as well ;-)
okay
Maybe implement something like this:
make import U_BOOT_TREE=/path/to/u-boot
I've hacked it together, along with
patches:
Acked-by: Remy Bohmer li...@bohmer.net
--
ptxdist mailing list
ptxdist@pengutronix.de
1 - 100 of 217 matches
Mail list logo