| Message ID | 1334149610-989-1-git-send-email-alexandre.belloni@piout.net |
|---|---|
| State | Accepted |
| Headers | show |
Le Wed, 11 Apr 2012 15:06:50 +0200, Alexandre Belloni <alexandre.belloni@piout.net> a écrit : > Signed-off-by: Alexandre Belloni <alexandre.belloni@piout.net> Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
On 15:35 Wed 11 Apr , Thomas Petazzoni wrote: > Le Wed, 11 Apr 2012 15:06:50 +0200, > Alexandre Belloni <alexandre.belloni@piout.net> a écrit : > > > Signed-off-by: Alexandre Belloni <alexandre.belloni@piout.net> > > Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> this soc config I disagree I run on ARM and PPC and samba is x86 only Best Regards, J.
On Wed, Apr 11, 2012 at 03:50:31PM +0200, Jean-Christophe PLAGNIOL-VILLARD wrote : > > Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> > this soc config I disagree > > I run on ARM and PPC and samba is x86 only > That makes sense. Should we detect the host arch and disable sam-ba ? I guess we can do that whitout overcomplicating the host infrastructure.
>>>>> "Jean-Christophe" == Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> writes:
Hi,
Jean-Christophe> this soc config I disagree
Jean-Christophe> I run on ARM and PPC and samba is x86 only
You build buildroot ON ARM/PPC? That's very uncommon (well, gcc110 is
ppc64 - But still). Nevertheless, nothing breaks by including sam-ba in
the build, your host dir just gets a little bit bigger. In this case I
would say the advantage for the common case (build on x86 for at91
development) outweighs the minor disadvantage for your use case.
>>>>> "Alexandre" == Alexandre Belloni <alexandre.belloni@piout.net> writes:
Alexandre> Signed-off-by: Alexandre Belloni <alexandre.belloni@piout.net>
Alexandre> ---
Alexandre> configs/at91sam9260dfc_defconfig | 4 ++++
Alexandre> configs/at91sam9261ek_defconfig | 5 +++++
Alexandre> configs/at91sam9263ek_defconfig | 4 ++++
Alexandre> 3 files changed, 13 insertions(+), 0 deletions(-)
You forgot 9g20dfc. I've fixed that and committed, thanks.
On 11:19 Sun 15 Apr , Peter Korsgaard wrote: > >>>>> "Jean-Christophe" == Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> writes: > > Hi, > > Jean-Christophe> this soc config I disagree > > Jean-Christophe> I run on ARM and PPC and samba is x86 only > > You build buildroot ON ARM/PPC? That's very uncommon (well, gcc110 is > ppc64 - But still). Nevertheless, nothing breaks by including sam-ba in > the build, your host dir just gets a little bit bigger. In this case I > would say the advantage for the common case (build on x86 for at91 > development) outweighs the minor disadvantage for your use case. I've a quad Arm and a quad G4 except you expect a packet you install to work I expect to detect the HOST type and propose it only if it work Best Regards, J.
>>>>> "Jean-Christophe" == Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> writes: Hi, >> You build buildroot ON ARM/PPC? That's very uncommon (well, gcc110 is >> ppc64 - But still). Nevertheless, nothing breaks by including sam-ba in >> the build, your host dir just gets a little bit bigger. In this case I >> would say the advantage for the common case (build on x86 for at91 >> development) outweighs the minor disadvantage for your use case. Jean-Christophe> I've a quad Arm and a quad G4 Both of those are significantly slower than a recent x86, but all the power to you for using it. Jean-Christophe> except you expect a packet you install to work I Jean-Christophe> expect to detect the HOST type and propose it only if Jean-Christophe> it work We don't have any concept of host type in Kconfig, so I don't see how we could possibly do that.
On 19:04 Sun 15 Apr , Peter Korsgaard wrote: > >>>>> "Jean-Christophe" == Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> writes: > > Hi, > > >> You build buildroot ON ARM/PPC? That's very uncommon (well, gcc110 is > >> ppc64 - But still). Nevertheless, nothing breaks by including sam-ba in > >> the build, your host dir just gets a little bit bigger. In this case I > >> would say the advantage for the common case (build on x86 for at91 > >> development) outweighs the minor disadvantage for your use case. > > Jean-Christophe> I've a quad Arm and a quad G4 > > Both of those are significantly slower than a recent x86, but all the > power to you for using it. > > Jean-Christophe> except you expect a packet you install to work I > Jean-Christophe> expect to detect the HOST type and propose it only if > Jean-Christophe> it work > > We don't have any concept of host type in Kconfig, so I don't see how we > could possibly do that. I did it on dtc HOSTOS := $(shell uname -s | tr '[:upper:]' '[:lower:]' | \ sed -e 's/\(cygwin\).*/cygwin/') this work on unix Best Regards, J.
Le Mon, 16 Apr 2012 14:34:29 +0200, Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> a écrit : > I did it on dtc > > HOSTOS := $(shell uname -s | tr '[:upper:]' '[:lower:]' | \ > sed -e 's/\(cygwin\).*/cygwin/') > > this work on unix But it does not allow to do Kconfig dependencies. Thomas
On 15:23 Mon 16 Apr , Thomas Petazzoni wrote: > Le Mon, 16 Apr 2012 14:34:29 +0200, > Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com> a écrit : > > > I did it on dtc > > > > HOSTOS := $(shell uname -s | tr '[:upper:]' '[:lower:]' | \ > > sed -e 's/\(cygwin\).*/cygwin/') > > > > this work on unix > > But it does not allow to do Kconfig dependencies. you can use env in Kconfig option env="KERNELVERSION" in barebox or kernel Best Regards, J.
diff --git a/configs/at91sam9260dfc_defconfig b/configs/at91sam9260dfc_defconfig index 3cd58ac..921afa8 100644 --- a/configs/at91sam9260dfc_defconfig +++ b/configs/at91sam9260dfc_defconfig @@ -2,6 +2,10 @@ BR2_arm=y BR2_arm926t=y +# Host utilities +# +BR2_PACKAGE_HOST_SAM_BA=y + # First stage bootloader BR2_TARGET_AT91BOOTSTRAP=y BR2_TARGET_AT91BOOTSTRAP_BOARD="at91sam9260ek" diff --git a/configs/at91sam9261ek_defconfig b/configs/at91sam9261ek_defconfig index 73edd89..b801bbe 100644 --- a/configs/at91sam9261ek_defconfig +++ b/configs/at91sam9261ek_defconfig @@ -2,6 +2,11 @@ BR2_arm=y BR2_arm926t=y + +# Host utilities +# +BR2_PACKAGE_HOST_SAM_BA=y + # First stage bootloader BR2_TARGET_AT91BOOTSTRAP=y BR2_TARGET_AT91BOOTSTRAP_BOARD="at91sam9261ek" diff --git a/configs/at91sam9263ek_defconfig b/configs/at91sam9263ek_defconfig index e4a44f1..2cd82c0 100644 --- a/configs/at91sam9263ek_defconfig +++ b/configs/at91sam9263ek_defconfig @@ -2,6 +2,10 @@ BR2_arm=y BR2_arm926t=y +# Host utilities +# +BR2_PACKAGE_HOST_SAM_BA=y + # First stage bootloader BR2_TARGET_AT91BOOTSTRAP=y BR2_TARGET_AT91BOOTSTRAP_BOARD="at91sam9263ek"
Signed-off-by: Alexandre Belloni <alexandre.belloni@piout.net> --- configs/at91sam9260dfc_defconfig | 4 ++++ configs/at91sam9261ek_defconfig | 5 +++++ configs/at91sam9263ek_defconfig | 4 ++++ 3 files changed, 13 insertions(+), 0 deletions(-)