Patchwork [1/2] lttng-modules: Update revision to grab last bugfix releases

login
register
mail settings
Submitter Otavio Salvador
Date Nov. 1, 2013, 1:31 p.m.
Message ID <1383312696-421-1-git-send-email-otavio@ossystems.com.br>
Download mbox | patch
Permalink /patch/60967/
State New
Headers show

Comments

Otavio Salvador - Nov. 1, 2013, 1:31 p.m.
This updates lttng-modules for 2.3.2 and also includes a bugfix
imported from 3.4 kernel. The changes included are:

6c26b01 Blacklist Linux kernels 3.10+
b97d469 LTTng trace-clock: shrink kernel blacklist
7d40f8e Version 2.3.1
2aecb9b Blacklist kernels 3.10.13 and 3.11.2
4f1217f Version 2.3.2
92fff0c Import fix from LTSI: 3.4+ RT kernels use CONFIG_PREEMPT_RT_FULL
1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty implementation

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
---
 meta/recipes-kernel/lttng/lttng-modules_git.bb | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
Saul Wold - Nov. 13, 2013, 9:54 p.m.
On 11/01/2013 06:31 AM, Otavio Salvador wrote:
> This updates lttng-modules for 2.3.2 and also includes a bugfix
> imported from 3.4 kernel. The changes included are:
>
> 6c26b01 Blacklist Linux kernels 3.10+
> b97d469 LTTng trace-clock: shrink kernel blacklist
> 7d40f8e Version 2.3.1
> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
> 4f1217f Version 2.3.2
> 92fff0c Import fix from LTSI: 3.4+ RT kernels use CONFIG_PREEMPT_RT_FULL
> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty implementation
>

Otavio,

Tried to build this locally and got the following failure with 
genericx86 and master:

|   CC [M] 
/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
| In file included from 
/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
|                  from 
/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
| 
/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2: 
error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the 
timekeeping subsystem. Fixed by commit 
7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK 
related deadlock from ntp lock changes\" in Linux."
|  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the 
timekeeping subsystem. Fixed by commit 
7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK 
related deadlock from ntp lock changes\" in Linux."
|   ^
| make[2]: *** 
[/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o] 
Error 1
| make[1]: *** 
[_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git] 
Error 2
| make[1]: Leaving directory 
`/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
| make: *** [default] Error 2
| ERROR: oe_runmake failed

So I am going to hold off on this patch set until you can resolve this.

Thanks
	Sau!

> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
> ---
>   meta/recipes-kernel/lttng/lttng-modules_git.bb | 4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/meta/recipes-kernel/lttng/lttng-modules_git.bb b/meta/recipes-kernel/lttng/lttng-modules_git.bb
> index 789b5ec..e887a86 100644
> --- a/meta/recipes-kernel/lttng/lttng-modules_git.bb
> +++ b/meta/recipes-kernel/lttng/lttng-modules_git.bb
> @@ -10,8 +10,8 @@ DEPENDS = "virtual/kernel"
>
>   inherit module
>
> -SRCREV = "78c8710bb8e6b7f6301d95afec0305f40fa063ad"
> -PV = "2.3.0+git${SRCPV}"
> +SRCREV = "18740750760fde894eac0cd4e3ef4d8c7c4ba8c3"
> +PV = "2.3.2+git${SRCPV}"
>
>   SRC_URI = "git://git.lttng.org/lttng-modules.git \
>              file://lttng-modules-replace-KERNELDIR-with-KERNEL_SRC.patch"
>
Otavio Salvador - Nov. 14, 2013, 1:45 a.m.
On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com> wrote:
> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>
>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>> imported from 3.4 kernel. The changes included are:
>>
>> 6c26b01 Blacklist Linux kernels 3.10+
>> b97d469 LTTng trace-clock: shrink kernel blacklist
>> 7d40f8e Version 2.3.1
>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>> 4f1217f Version 2.3.2
>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use CONFIG_PREEMPT_RT_FULL
>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>> implementation
>>
>
> Otavio,
>
> Tried to build this locally and got the following failure with genericx86
> and master:
>
> |   CC [M]
> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
> | In file included from
> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
> |                  from
> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
> |
> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
> timekeeping subsystem. Fixed by commit
> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK related
> deadlock from ntp lock changes\" in Linux."
> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
> timekeeping subsystem. Fixed by commit
> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK related
> deadlock from ntp lock changes\" in Linux."
> |   ^
> | make[2]: ***
> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
> Error 1
> | make[1]: ***
> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
> Error 2
> | make[1]: Leaving directory
> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
> | make: *** [default] Error 2
> | ERROR: oe_runmake failed
>
> So I am going to hold off on this patch set until you can resolve this.

Yes; this is becase the linux-yocto seems to not have been updated to
> 3.10.13; so in fact it is failing to build to avoid people to
trigger the deadlock.

The change which introduces this in lttng-modules is:

http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296

I can update this patch as 2.3.3 has been released and it includes the
3.12 fixes but please check when linux-yocto will be updated so we can
fix it.

Regards,
Bruce Ashfield - Nov. 14, 2013, 3:08 a.m.
On Wed, Nov 13, 2013 at 8:45 PM, Otavio Salvador
<otavio@ossystems.com.br> wrote:
> On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com> wrote:
>> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>>
>>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>>> imported from 3.4 kernel. The changes included are:
>>>
>>> 6c26b01 Blacklist Linux kernels 3.10+
>>> b97d469 LTTng trace-clock: shrink kernel blacklist
>>> 7d40f8e Version 2.3.1
>>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>>> 4f1217f Version 2.3.2
>>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use CONFIG_PREEMPT_RT_FULL
>>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>>> implementation
>>>
>>
>> Otavio,
>>
>> Tried to build this locally and got the following failure with genericx86
>> and master:
>>
>> |   CC [M]
>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
>> | In file included from
>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
>> |                  from
>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
>> |
>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
>> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>> timekeeping subsystem. Fixed by commit
>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK related
>> deadlock from ntp lock changes\" in Linux."
>> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>> timekeeping subsystem. Fixed by commit
>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK related
>> deadlock from ntp lock changes\" in Linux."
>> |   ^
>> | make[2]: ***
>> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
>> Error 1
>> | make[1]: ***
>> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
>> Error 2
>> | make[1]: Leaving directory
>> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
>> | make: *** [default] Error 2
>> | ERROR: oe_runmake failed
>>
>> So I am going to hold off on this patch set until you can resolve this.
>
> Yes; this is becase the linux-yocto seems to not have been updated to
>> 3.10.13; so in fact it is failing to build to avoid people to
> trigger the deadlock.
>
> The change which introduces this in lttng-modules is:
>
> http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296
>
> I can update this patch as 2.3.3 has been released and it includes the
> 3.12 fixes but please check when linux-yocto will be updated so we can
> fix it.

Saul: What board are you building ? I've had linux-yocto on 3.10.17
for over a week
now.

The meta-yocto-bsps need a SRCREV bump, which I was just about to send
out, but the code has been in tree for some time yet, and all the qemu*
BSPs are already running at that revision.

Bruce

>
> Regards,
>
>
> --
> Otavio Salvador                             O.S. Systems
> http://www.ossystems.com.br        http://code.ossystems.com.br
> Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
Saul Wold - Nov. 14, 2013, 4:09 p.m.
On 11/13/2013 07:08 PM, Bruce Ashfield wrote:
> On Wed, Nov 13, 2013 at 8:45 PM, Otavio Salvador
> <otavio@ossystems.com.br> wrote:
>> On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com> wrote:
>>> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>>>
>>>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>>>> imported from 3.4 kernel. The changes included are:
>>>>
>>>> 6c26b01 Blacklist Linux kernels 3.10+
>>>> b97d469 LTTng trace-clock: shrink kernel blacklist
>>>> 7d40f8e Version 2.3.1
>>>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>>>> 4f1217f Version 2.3.2
>>>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use CONFIG_PREEMPT_RT_FULL
>>>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>>>> implementation
>>>>
>>>
>>> Otavio,
>>>
>>> Tried to build this locally and got the following failure with genericx86
>>> and master:
>>>
>>> |   CC [M]
>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
>>> | In file included from
>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
>>> |                  from
>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
>>> |
>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
>>> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>> timekeeping subsystem. Fixed by commit
>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK related
>>> deadlock from ntp lock changes\" in Linux."
>>> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>> timekeeping subsystem. Fixed by commit
>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK related
>>> deadlock from ntp lock changes\" in Linux."
>>> |   ^
>>> | make[2]: ***
>>> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
>>> Error 1
>>> | make[1]: ***
>>> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
>>> Error 2
>>> | make[1]: Leaving directory
>>> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
>>> | make: *** [default] Error 2
>>> | ERROR: oe_runmake failed
>>>
>>> So I am going to hold off on this patch set until you can resolve this.
>>
>> Yes; this is becase the linux-yocto seems to not have been updated to
>>> 3.10.13; so in fact it is failing to build to avoid people to
>> trigger the deadlock.
>>
>> The change which introduces this in lttng-modules is:
>>
>> http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296
>>
>> I can update this patch as 2.3.3 has been released and it includes the
>> 3.12 fixes but please check when linux-yocto will be updated so we can
>> fix it.
>
> Saul: What board are you building ? I've had linux-yocto on 3.10.17
> for over a week
> now.

As mentioned above this is for genericx86, so maybe not all the BSPs are 
updated yet.

Sau!

>
> The meta-yocto-bsps need a SRCREV bump, which I was just about to send
> out, but the code has been in tree for some time yet, and all the qemu*
> BSPs are already running at that revision.
>
> Bruce
>
>>
>> Regards,
>>
>>
>> --
>> Otavio Salvador                             O.S. Systems
>> http://www.ossystems.com.br        http://code.ossystems.com.br
>> Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750
>> _______________________________________________
>> Openembedded-core mailing list
>> Openembedded-core@lists.openembedded.org
>> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>
>
>
Bruce Ashfield - Nov. 14, 2013, 4:18 p.m.
On Thu, Nov 14, 2013 at 11:09 AM, Saul Wold <sgw@linux.intel.com> wrote:
> On 11/13/2013 07:08 PM, Bruce Ashfield wrote:
>>
>> On Wed, Nov 13, 2013 at 8:45 PM, Otavio Salvador
>> <otavio@ossystems.com.br> wrote:
>>>
>>> On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com> wrote:
>>>>
>>>> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>>>>
>>>>>
>>>>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>>>>> imported from 3.4 kernel. The changes included are:
>>>>>
>>>>> 6c26b01 Blacklist Linux kernels 3.10+
>>>>> b97d469 LTTng trace-clock: shrink kernel blacklist
>>>>> 7d40f8e Version 2.3.1
>>>>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>>>>> 4f1217f Version 2.3.2
>>>>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use
>>>>> CONFIG_PREEMPT_RT_FULL
>>>>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>>>>> implementation
>>>>>
>>>>
>>>> Otavio,
>>>>
>>>> Tried to build this locally and got the following failure with
>>>> genericx86
>>>> and master:
>>>>
>>>> |   CC [M]
>>>>
>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
>>>> | In file included from
>>>>
>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
>>>> |                  from
>>>>
>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
>>>> |
>>>>
>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
>>>> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>> timekeeping subsystem. Fixed by commit
>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>> related
>>>> deadlock from ntp lock changes\" in Linux."
>>>> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>> timekeeping subsystem. Fixed by commit
>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>> related
>>>> deadlock from ntp lock changes\" in Linux."
>>>> |   ^
>>>> | make[2]: ***
>>>>
>>>> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
>>>> Error 1
>>>> | make[1]: ***
>>>>
>>>> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
>>>> Error 2
>>>> | make[1]: Leaving directory
>>>> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
>>>> | make: *** [default] Error 2
>>>> | ERROR: oe_runmake failed
>>>>
>>>> So I am going to hold off on this patch set until you can resolve this.
>>>
>>>
>>> Yes; this is becase the linux-yocto seems to not have been updated to
>>>>
>>>> 3.10.13; so in fact it is failing to build to avoid people to
>>>
>>> trigger the deadlock.
>>>
>>> The change which introduces this in lttng-modules is:
>>>
>>>
>>> http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296
>>>
>>> I can update this patch as 2.3.3 has been released and it includes the
>>> 3.12 fixes but please check when linux-yocto will be updated so we can
>>> fix it.
>>
>>
>> Saul: What board are you building ? I've had linux-yocto on 3.10.17
>> for over a week
>> now.
>
>
> As mentioned above this is for genericx86, so maybe not all the BSPs are
> updated yet.

Buried in the context, sorry about missing that.

Yes, I'm about to update the reference boards, but there's a install
error that I'm fixing first, but I'll have the series out by end of day.

Bruce

>
> Sau!
>
>
>>
>> The meta-yocto-bsps need a SRCREV bump, which I was just about to send
>> out, but the code has been in tree for some time yet, and all the qemu*
>> BSPs are already running at that revision.
>>
>> Bruce
>>
>>>
>>> Regards,
>>>
>>>
>>> --
>>> Otavio Salvador                             O.S. Systems
>>> http://www.ossystems.com.br        http://code.ossystems.com.br
>>> Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750
>>> _______________________________________________
>>> Openembedded-core mailing list
>>> Openembedded-core@lists.openembedded.org
>>> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>>
>>
>>
>>
>
Bruce Ashfield - Nov. 14, 2013, 5:15 p.m.
On Thu, Nov 14, 2013 at 11:18 AM, Bruce Ashfield
<bruce.ashfield@gmail.com> wrote:
> On Thu, Nov 14, 2013 at 11:09 AM, Saul Wold <sgw@linux.intel.com> wrote:
>> On 11/13/2013 07:08 PM, Bruce Ashfield wrote:
>>>
>>> On Wed, Nov 13, 2013 at 8:45 PM, Otavio Salvador
>>> <otavio@ossystems.com.br> wrote:
>>>>
>>>> On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com> wrote:
>>>>>
>>>>> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>>>>>
>>>>>>
>>>>>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>>>>>> imported from 3.4 kernel. The changes included are:
>>>>>>
>>>>>> 6c26b01 Blacklist Linux kernels 3.10+
>>>>>> b97d469 LTTng trace-clock: shrink kernel blacklist
>>>>>> 7d40f8e Version 2.3.1
>>>>>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>>>>>> 4f1217f Version 2.3.2
>>>>>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use
>>>>>> CONFIG_PREEMPT_RT_FULL
>>>>>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>>>>>> implementation
>>>>>>
>>>>>
>>>>> Otavio,
>>>>>
>>>>> Tried to build this locally and got the following failure with
>>>>> genericx86
>>>>> and master:
>>>>>
>>>>> |   CC [M]
>>>>>
>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
>>>>> | In file included from
>>>>>
>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
>>>>> |                  from
>>>>>
>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
>>>>> |
>>>>>
>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
>>>>> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>>> timekeeping subsystem. Fixed by commit
>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>> related
>>>>> deadlock from ntp lock changes\" in Linux."
>>>>> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>>> timekeeping subsystem. Fixed by commit
>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>> related
>>>>> deadlock from ntp lock changes\" in Linux."
>>>>> |   ^
>>>>> | make[2]: ***
>>>>>
>>>>> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
>>>>> Error 1
>>>>> | make[1]: ***
>>>>>
>>>>> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
>>>>> Error 2
>>>>> | make[1]: Leaving directory
>>>>> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
>>>>> | make: *** [default] Error 2
>>>>> | ERROR: oe_runmake failed
>>>>>
>>>>> So I am going to hold off on this patch set until you can resolve this.
>>>>
>>>>
>>>> Yes; this is becase the linux-yocto seems to not have been updated to
>>>>>
>>>>> 3.10.13; so in fact it is failing to build to avoid people to
>>>>
>>>> trigger the deadlock.
>>>>
>>>> The change which introduces this in lttng-modules is:
>>>>
>>>>
>>>> http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296
>>>>
>>>> I can update this patch as 2.3.3 has been released and it includes the
>>>> 3.12 fixes but please check when linux-yocto will be updated so we can
>>>> fix it.
>>>
>>>
>>> Saul: What board are you building ? I've had linux-yocto on 3.10.17
>>> for over a week
>>> now.
>>
>>
>> As mentioned above this is for genericx86, so maybe not all the BSPs are
>> updated yet.
>
> Buried in the context, sorry about missing that.
>
> Yes, I'm about to update the reference boards, but there's a install
> error that I'm fixing first, but I'll have the series out by end of day.

I just sent the update for this (cc'd poky@yoctoproject.org), but I haven't
seen it loop around.

If you didn't get the patch, let me know and I'll try again.

Bruce

>
> Bruce
>
>>
>> Sau!
>>
>>
>>>
>>> The meta-yocto-bsps need a SRCREV bump, which I was just about to send
>>> out, but the code has been in tree for some time yet, and all the qemu*
>>> BSPs are already running at that revision.
>>>
>>> Bruce
>>>
>>>>
>>>> Regards,
>>>>
>>>>
>>>> --
>>>> Otavio Salvador                             O.S. Systems
>>>> http://www.ossystems.com.br        http://code.ossystems.com.br
>>>> Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750
>>>> _______________________________________________
>>>> Openembedded-core mailing list
>>>> Openembedded-core@lists.openembedded.org
>>>> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>>>
>>>
>>>
>>>
>>
>
>
>
> --
> "Thou shalt not follow the NULL pointer, for chaos and madness await
> thee at its end"
Otavio Salvador - Nov. 14, 2013, 5:48 p.m.
On Thu, Nov 14, 2013 at 3:15 PM, Bruce Ashfield
<bruce.ashfield@gmail.com> wrote:
> On Thu, Nov 14, 2013 at 11:18 AM, Bruce Ashfield
> <bruce.ashfield@gmail.com> wrote:
>> On Thu, Nov 14, 2013 at 11:09 AM, Saul Wold <sgw@linux.intel.com> wrote:
>>> On 11/13/2013 07:08 PM, Bruce Ashfield wrote:
>>>>
>>>> On Wed, Nov 13, 2013 at 8:45 PM, Otavio Salvador
>>>> <otavio@ossystems.com.br> wrote:
>>>>>
>>>>> On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com> wrote:
>>>>>>
>>>>>> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>>>>>>
>>>>>>>
>>>>>>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>>>>>>> imported from 3.4 kernel. The changes included are:
>>>>>>>
>>>>>>> 6c26b01 Blacklist Linux kernels 3.10+
>>>>>>> b97d469 LTTng trace-clock: shrink kernel blacklist
>>>>>>> 7d40f8e Version 2.3.1
>>>>>>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>>>>>>> 4f1217f Version 2.3.2
>>>>>>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use
>>>>>>> CONFIG_PREEMPT_RT_FULL
>>>>>>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>>>>>>> implementation
>>>>>>>
>>>>>>
>>>>>> Otavio,
>>>>>>
>>>>>> Tried to build this locally and got the following failure with
>>>>>> genericx86
>>>>>> and master:
>>>>>>
>>>>>> |   CC [M]
>>>>>>
>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
>>>>>> | In file included from
>>>>>>
>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
>>>>>> |                  from
>>>>>>
>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
>>>>>> |
>>>>>>
>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
>>>>>> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>>>> timekeeping subsystem. Fixed by commit
>>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>>> related
>>>>>> deadlock from ntp lock changes\" in Linux."
>>>>>> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>>>> timekeeping subsystem. Fixed by commit
>>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>>> related
>>>>>> deadlock from ntp lock changes\" in Linux."
>>>>>> |   ^
>>>>>> | make[2]: ***
>>>>>>
>>>>>> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
>>>>>> Error 1
>>>>>> | make[1]: ***
>>>>>>
>>>>>> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
>>>>>> Error 2
>>>>>> | make[1]: Leaving directory
>>>>>> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
>>>>>> | make: *** [default] Error 2
>>>>>> | ERROR: oe_runmake failed
>>>>>>
>>>>>> So I am going to hold off on this patch set until you can resolve this.
>>>>>
>>>>>
>>>>> Yes; this is becase the linux-yocto seems to not have been updated to
>>>>>>
>>>>>> 3.10.13; so in fact it is failing to build to avoid people to
>>>>>
>>>>> trigger the deadlock.
>>>>>
>>>>> The change which introduces this in lttng-modules is:
>>>>>
>>>>>
>>>>> http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296
>>>>>
>>>>> I can update this patch as 2.3.3 has been released and it includes the
>>>>> 3.12 fixes but please check when linux-yocto will be updated so we can
>>>>> fix it.
>>>>
>>>>
>>>> Saul: What board are you building ? I've had linux-yocto on 3.10.17
>>>> for over a week
>>>> now.
>>>
>>>
>>> As mentioned above this is for genericx86, so maybe not all the BSPs are
>>> updated yet.
>>
>> Buried in the context, sorry about missing that.
>>
>> Yes, I'm about to update the reference boards, but there's a install
>> error that I'm fixing first, but I'll have the series out by end of day.
>
> I just sent the update for this (cc'd poky@yoctoproject.org), but I haven't
> seen it loop around.
>
> If you didn't get the patch, let me know and I'll try again.

I have sent the new patch version.

It does the update to the 2.3.3 which include all the previous
backported patches.

This should be considered for Dora as well as it fixes the deadlock
and building with 3.12 kernels.
Robert Yang - Nov. 18, 2013, 12:25 p.m.
On 11/15/2013 01:48 AM, Otavio Salvador wrote:
> On Thu, Nov 14, 2013 at 3:15 PM, Bruce Ashfield
> <bruce.ashfield@gmail.com> wrote:
>> On Thu, Nov 14, 2013 at 11:18 AM, Bruce Ashfield
>> <bruce.ashfield@gmail.com> wrote:
>>> On Thu, Nov 14, 2013 at 11:09 AM, Saul Wold <sgw@linux.intel.com> wrote:
>>>> On 11/13/2013 07:08 PM, Bruce Ashfield wrote:
>>>>>
>>>>> On Wed, Nov 13, 2013 at 8:45 PM, Otavio Salvador
>>>>> <otavio@ossystems.com.br> wrote:
>>>>>>
>>>>>> On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com> wrote:
>>>>>>>
>>>>>>> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>>>>>>>> imported from 3.4 kernel. The changes included are:
>>>>>>>>
>>>>>>>> 6c26b01 Blacklist Linux kernels 3.10+
>>>>>>>> b97d469 LTTng trace-clock: shrink kernel blacklist
>>>>>>>> 7d40f8e Version 2.3.1
>>>>>>>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>>>>>>>> 4f1217f Version 2.3.2
>>>>>>>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use
>>>>>>>> CONFIG_PREEMPT_RT_FULL
>>>>>>>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>>>>>>>> implementation
>>>>>>>>
>>>>>>>
>>>>>>> Otavio,
>>>>>>>
>>>>>>> Tried to build this locally and got the following failure with
>>>>>>> genericx86
>>>>>>> and master:
>>>>>>>
>>>>>>> |   CC [M]
>>>>>>>
>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
>>>>>>> | In file included from
>>>>>>>
>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
>>>>>>> |                  from
>>>>>>>
>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
>>>>>>> |
>>>>>>>
>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
>>>>>>> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>>>>> timekeeping subsystem. Fixed by commit
>>>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>>>> related
>>>>>>> deadlock from ntp lock changes\" in Linux."
>>>>>>> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>>>>> timekeeping subsystem. Fixed by commit
>>>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>>>> related
>>>>>>> deadlock from ntp lock changes\" in Linux."
>>>>>>> |   ^
>>>>>>> | make[2]: ***
>>>>>>>
>>>>>>> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
>>>>>>> Error 1
>>>>>>> | make[1]: ***
>>>>>>>
>>>>>>> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
>>>>>>> Error 2
>>>>>>> | make[1]: Leaving directory
>>>>>>> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
>>>>>>> | make: *** [default] Error 2
>>>>>>> | ERROR: oe_runmake failed
>>>>>>>
>>>>>>> So I am going to hold off on this patch set until you can resolve this.
>>>>>>
>>>>>>
>>>>>> Yes; this is becase the linux-yocto seems to not have been updated to
>>>>>>>
>>>>>>> 3.10.13; so in fact it is failing to build to avoid people to
>>>>>>
>>>>>> trigger the deadlock.
>>>>>>
>>>>>> The change which introduces this in lttng-modules is:
>>>>>>
>>>>>>
>>>>>> http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296
>>>>>>
>>>>>> I can update this patch as 2.3.3 has been released and it includes the
>>>>>> 3.12 fixes but please check when linux-yocto will be updated so we can
>>>>>> fix it.
>>>>>
>>>>>
>>>>> Saul: What board are you building ? I've had linux-yocto on 3.10.17
>>>>> for over a week
>>>>> now.
>>>>
>>>>
>>>> As mentioned above this is for genericx86, so maybe not all the BSPs are
>>>> updated yet.
>>>
>>> Buried in the context, sorry about missing that.
>>>
>>> Yes, I'm about to update the reference boards, but there's a install
>>> error that I'm fixing first, but I'll have the series out by end of day.
>>
>> I just sent the update for this (cc'd poky@yoctoproject.org), but I haven't
>> seen it loop around.
>>
>> If you didn't get the patch, let me know and I'll try again.
>
> I have sent the new patch version.
>
> It does the update to the 2.3.3 which include all the previous
> backported patches.
>
> This should be considered for Dora as well as it fixes the deadlock
> and building with 3.12 kernels.
>

Hi Otavio, thanks for suggesting this for dora, but dora's kernel is 3.10.11,
and this is an update for lttng-modules, I'm not sure whether dora needs this.

Add Saul in the CC list.

// Robert
Otavio Salvador - Nov. 18, 2013, 2:18 p.m.
On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang <liezhi.yang@windriver.com> wrote:
>
>
> On 11/15/2013 01:48 AM, Otavio Salvador wrote:
>>
>> On Thu, Nov 14, 2013 at 3:15 PM, Bruce Ashfield
>> <bruce.ashfield@gmail.com> wrote:
>>>
>>> On Thu, Nov 14, 2013 at 11:18 AM, Bruce Ashfield
>>> <bruce.ashfield@gmail.com> wrote:
>>>>
>>>> On Thu, Nov 14, 2013 at 11:09 AM, Saul Wold <sgw@linux.intel.com> wrote:
>>>>>
>>>>> On 11/13/2013 07:08 PM, Bruce Ashfield wrote:
>>>>>>
>>>>>>
>>>>>> On Wed, Nov 13, 2013 at 8:45 PM, Otavio Salvador
>>>>>> <otavio@ossystems.com.br> wrote:
>>>>>>>
>>>>>>>
>>>>>>> On Wed, Nov 13, 2013 at 7:54 PM, Saul Wold <sgw@linux.intel.com>
>>>>>>> wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>> On 11/01/2013 06:31 AM, Otavio Salvador wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> This updates lttng-modules for 2.3.2 and also includes a bugfix
>>>>>>>>> imported from 3.4 kernel. The changes included are:
>>>>>>>>>
>>>>>>>>> 6c26b01 Blacklist Linux kernels 3.10+
>>>>>>>>> b97d469 LTTng trace-clock: shrink kernel blacklist
>>>>>>>>> 7d40f8e Version 2.3.1
>>>>>>>>> 2aecb9b Blacklist kernels 3.10.13 and 3.11.2
>>>>>>>>> 4f1217f Version 2.3.2
>>>>>>>>> 92fff0c Import fix from LTSI: 3.4+ RT kernels use
>>>>>>>>> CONFIG_PREEMPT_RT_FULL
>>>>>>>>> 1874075 lttng-statedump-impl: Fix lttng_list_interrupts for empty
>>>>>>>>> implementation
>>>>>>>>>
>>>>>>>>
>>>>>>>> Otavio,
>>>>>>>>
>>>>>>>> Tried to build this locally and got the following failure with
>>>>>>>> genericx86
>>>>>>>> and master:
>>>>>>>>
>>>>>>>> |   CC [M]
>>>>>>>>
>>>>>>>>
>>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o
>>>>>>>> | In file included from
>>>>>>>>
>>>>>>>>
>>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-tracer.h:37:0,
>>>>>>>> |                  from
>>>>>>>>
>>>>>>>>
>>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.c:24:
>>>>>>>> |
>>>>>>>>
>>>>>>>>
>>>>>>>> /srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/wrapper/trace-clock.h:40:2:
>>>>>>>> error: #error "Linux kernels 3.10 and 3.11 introduce a deadlock in
>>>>>>>> the
>>>>>>>> timekeeping subsystem. Fixed by commit
>>>>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>>>>> related
>>>>>>>> deadlock from ntp lock changes\" in Linux."
>>>>>>>> |  #error "Linux kernels 3.10 and 3.11 introduce a deadlock in the
>>>>>>>> timekeeping subsystem. Fixed by commit
>>>>>>>> 7bd36014460f793c19e7d6c94dab67b0afcfcb7f \"timekeeping: Fix HRTICK
>>>>>>>> related
>>>>>>>> deadlock from ntp lock changes\" in Linux."
>>>>>>>> |   ^
>>>>>>>> | make[2]: ***
>>>>>>>>
>>>>>>>>
>>>>>>>> [/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git/lttng-ring-buffer-client-discard.o]
>>>>>>>> Error 1
>>>>>>>> | make[1]: ***
>>>>>>>>
>>>>>>>>
>>>>>>>> [_module_/srv/ssd/sgw/builds/world/tmp/work/genericx86-poky-linux/lttng-modules/2.3.2+gitAUTOINC+1874075076-r0/git]
>>>>>>>> Error 2
>>>>>>>> | make[1]: Leaving directory
>>>>>>>> `/srv/ssd/sgw/builds/world/tmp/sysroots/genericx86/usr/src/kernel'
>>>>>>>> | make: *** [default] Error 2
>>>>>>>> | ERROR: oe_runmake failed
>>>>>>>>
>>>>>>>> So I am going to hold off on this patch set until you can resolve
>>>>>>>> this.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Yes; this is becase the linux-yocto seems to not have been updated to
>>>>>>>>
>>>>>>>>
>>>>>>>> 3.10.13; so in fact it is failing to build to avoid people to
>>>>>>>
>>>>>>>
>>>>>>> trigger the deadlock.
>>>>>>>
>>>>>>> The change which introduces this in lttng-modules is:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> http://git.lttng.org/?p=lttng-modules.git;a=commit;h=2aecb9b1cbb63f6c524ad9195d8fd06299ef9296
>>>>>>>
>>>>>>> I can update this patch as 2.3.3 has been released and it includes
>>>>>>> the
>>>>>>> 3.12 fixes but please check when linux-yocto will be updated so we
>>>>>>> can
>>>>>>> fix it.
>>>>>>
>>>>>>
>>>>>>
>>>>>> Saul: What board are you building ? I've had linux-yocto on 3.10.17
>>>>>> for over a week
>>>>>> now.
>>>>>
>>>>>
>>>>>
>>>>> As mentioned above this is for genericx86, so maybe not all the BSPs
>>>>> are
>>>>> updated yet.
>>>>
>>>>
>>>> Buried in the context, sorry about missing that.
>>>>
>>>> Yes, I'm about to update the reference boards, but there's a install
>>>> error that I'm fixing first, but I'll have the series out by end of day.
>>>
>>>
>>> I just sent the update for this (cc'd poky@yoctoproject.org), but I
>>> haven't
>>> seen it loop around.
>>>
>>> If you didn't get the patch, let me know and I'll try again.
>>
>>
>> I have sent the new patch version.
>>
>> It does the update to the 2.3.3 which include all the previous
>> backported patches.
>>
>> This should be considered for Dora as well as it fixes the deadlock
>> and building with 3.12 kernels.
>>
>
> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
> 3.10.11,
> and this is an update for lttng-modules, I'm not sure whether dora needs
> this.
>
> Add Saul in the CC list.

I think Dora ought to get the new 3.10 kernel updates; this can go
along side with it.
Robert Yang - Nov. 18, 2013, 2:50 p.m.
On 11/18/2013 10:18 PM, Otavio Salvador wrote:
> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang <liezhi.yang@windriver.com> wrote:
>>>
>>> This should be considered for Dora as well as it fixes the deadlock
>>> and building with 3.12 kernels.
>>>
>>
>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
>> 3.10.11,
>> and this is an update for lttng-modules, I'm not sure whether dora needs
>> this.
>>
>> Add Saul in the CC list.
>
> I think Dora ought to get the new 3.10 kernel updates; this can go
> along side with it.
>

Hi Otavio,

I'm sorry, but I'm not sure whether we need update dora's kernel from 3.10.11
to 3.10.17, since we seldom do the package update for a stable branch,
let's wait for others' comments:-).

// Robert
Saul Wold - Nov. 18, 2013, 3:31 p.m.
On 11/18/2013 06:50 AM, Robert Yang wrote:
>
> On 11/18/2013 10:18 PM, Otavio Salvador wrote:
>> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
>> <liezhi.yang@windriver.com> wrote:
>>>>
>>>> This should be considered for Dora as well as it fixes the deadlock
>>>> and building with 3.12 kernels.
>>>>
>>>
>>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
>>> 3.10.11,
>>> and this is an update for lttng-modules, I'm not sure whether dora needs
>>> this.
>>>
>>> Add Saul in the CC list.
>>
>> I think Dora ought to get the new 3.10 kernel updates; this can go
>> along side with it.
>>
>
> Hi Otavio,
>
> I'm sorry, but I'm not sure whether we need update dora's kernel from
> 3.10.11
> to 3.10.17, since we seldom do the package update for a stable branch,
> let's wait for others' comments:-).
>
This is correct, I do not think we will be updating the kernel for a 
stable branch.

Sau!

> // Robert
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>
>
Otavio Salvador - Nov. 18, 2013, 3:50 p.m.
On Mon, Nov 18, 2013 at 1:31 PM, Saul Wold <saul.wold@intel.com> wrote:
> On 11/18/2013 06:50 AM, Robert Yang wrote:
>>
>>
>> On 11/18/2013 10:18 PM, Otavio Salvador wrote:
>>>
>>> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
>>> <liezhi.yang@windriver.com> wrote:
>>>>>
>>>>>
>>>>> This should be considered for Dora as well as it fixes the deadlock
>>>>> and building with 3.12 kernels.
>>>>>
>>>>
>>>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
>>>> 3.10.11,
>>>> and this is an update for lttng-modules, I'm not sure whether dora needs
>>>> this.
>>>>
>>>> Add Saul in the CC list.
>>>
>>>
>>> I think Dora ought to get the new 3.10 kernel updates; this can go
>>> along side with it.
>>>
>>
>> Hi Otavio,
>>
>> I'm sorry, but I'm not sure whether we need update dora's kernel from
>> 3.10.11
>> to 3.10.17, since we seldom do the package update for a stable branch,
>> let's wait for others' comments:-).
>>
> This is correct, I do not think we will be updating the kernel for a stable
> branch.

This is wrong. We should update to 3.10.17 as this is the LTS release,
otherwise what is the point of us using a LTS at all?
Paul Eggleton - Nov. 18, 2013, 3:56 p.m.
On Monday 18 November 2013 07:31:16 Saul Wold wrote:
> On 11/18/2013 06:50 AM, Robert Yang wrote:
> > On 11/18/2013 10:18 PM, Otavio Salvador wrote:
> >> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
> >> 
> >> <liezhi.yang@windriver.com> wrote:
> >>>> This should be considered for Dora as well as it fixes the deadlock
> >>>> and building with 3.12 kernels.
> >>> 
> >>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
> >>> 3.10.11,
> >>> and this is an update for lttng-modules, I'm not sure whether dora needs
> >>> this.
> >>> 
> >>> Add Saul in the CC list.
> >> 
> >> I think Dora ought to get the new 3.10 kernel updates; this can go
> >> along side with it.
> > 
> > Hi Otavio,
> > 
> > I'm sorry, but I'm not sure whether we need update dora's kernel from
> > 3.10.11
> > to 3.10.17, since we seldom do the package update for a stable branch,
> > let's wait for others' comments:-).
> 
> This is correct, I do not think we will be updating the kernel for a
> stable branch.

FWIW, there is a precedent for this in dylan - I relied upon Bruce sending 
patches that he felt were appropriate for the branch, and he did that on at 
least one occasion.

Cheers,
Paul
Bruce Ashfield - Nov. 18, 2013, 3:59 p.m.
On 13-11-18 10:50 AM, Otavio Salvador wrote:
> On Mon, Nov 18, 2013 at 1:31 PM, Saul Wold <saul.wold@intel.com> wrote:
>> On 11/18/2013 06:50 AM, Robert Yang wrote:
>>>
>>>
>>> On 11/18/2013 10:18 PM, Otavio Salvador wrote:
>>>>
>>>> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
>>>> <liezhi.yang@windriver.com> wrote:
>>>>>>
>>>>>>
>>>>>> This should be considered for Dora as well as it fixes the deadlock
>>>>>> and building with 3.12 kernels.
>>>>>>
>>>>>
>>>>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
>>>>> 3.10.11,
>>>>> and this is an update for lttng-modules, I'm not sure whether dora needs
>>>>> this.
>>>>>
>>>>> Add Saul in the CC list.
>>>>
>>>>
>>>> I think Dora ought to get the new 3.10 kernel updates; this can go
>>>> along side with it.
>>>>
>>>
>>> Hi Otavio,
>>>
>>> I'm sorry, but I'm not sure whether we need update dora's kernel from
>>> 3.10.11
>>> to 3.10.17, since we seldom do the package update for a stable branch,
>>> let's wait for others' comments:-).
>>>
>> This is correct, I do not think we will be updating the kernel for a stable
>> branch.
>
> This is wrong. We should update to 3.10.17 as this is the LTS release,
> otherwise what is the point of us using a LTS at all?

Obviously I'll be maintaining the LTS release in master, but that
maintenance gets a mix of bug fixes and features associated with the
upstream maintenance + LTSI + BSP requirements.

I've sent patches for the sustained branches in the past, and can
do it here as well. But to do it properly, there really needs to be a
dedicated kernel repository, since things like the LTSI import and
new BSP support will be mixed into the tree if a single 3.10 repo
is used.

I'm willing to split the repo and do -stable korg release updates, but
at some point there are only so many repositories and versions that
should be maintained at one time. As to what that number is, I don't have
a firm one in mind.

At a minimum, it would be safe for the dora maintenance branch to take
my latest kernel -stable imports, but in the not to distant future,
it would need a separate maintenance stream.

Cheers,

Bruce

>
Robert Yang - Nov. 19, 2013, 3:34 a.m.
Hi Bruce,

Here is the log status of the kernel layer between dora and master (18 commits).

I wonder are there any serious problems that we must upgrade it ?

$ git log dora..master --abbrev-commit --pretty=oneline meta/recipes-kernel

9a9a597 lttng-tools: Fixes incorrect path of ptest cases
e1e1ee5 linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for Beagleboard
779dec9 linux-yocto-rt/3.10: fix ntp merge issue
89c986c linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for Beagleboard
8a03d59 sysvinit: adjust boot sequence and remove hack from udev
e9ec153 linux-yocto/3.10: fix qemuarm boot and spurious mips build warning
baf005b linux-yocto/3.10: common-pc: add missing dependencies for BRCMSMAC
dd361f0 linux-yocto/3.8: add crystalforest bsp legacy block drivers configurations
1fe7bca linux-yocto/3.10: haswell-wc and crystalforest support
a23c7e9 linux-yocto-3.10: bump to 3.10.17 and -rt11
3129619 lttng-tools: make ptest able to work on target
5e42796 recipes: Remove PR = r0 from all recipes
2570f13 perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for mips64
56cc2f2 linux-yocto/3.10: MinnowBoard support
db1d37a perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for powerpc64
cd414c0 kmod: Update to Rev 15 via git
e83640d kmod: Add patch to fix seperate build dir of ptest
0f431a3 kmod-native: use bswap to work on older Linux hosts

// Robert

On 11/18/2013 11:59 PM, Bruce Ashfield wrote:
> On 13-11-18 10:50 AM, Otavio Salvador wrote:
>> On Mon, Nov 18, 2013 at 1:31 PM, Saul Wold <saul.wold@intel.com> wrote:
>>> On 11/18/2013 06:50 AM, Robert Yang wrote:
>>>>
>>>>
>>>> On 11/18/2013 10:18 PM, Otavio Salvador wrote:
>>>>>
>>>>> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
>>>>> <liezhi.yang@windriver.com> wrote:
>>>>>>>
>>>>>>>
>>>>>>> This should be considered for Dora as well as it fixes the deadlock
>>>>>>> and building with 3.12 kernels.
>>>>>>>
>>>>>>
>>>>>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
>>>>>> 3.10.11,
>>>>>> and this is an update for lttng-modules, I'm not sure whether dora needs
>>>>>> this.
>>>>>>
>>>>>> Add Saul in the CC list.
>>>>>
>>>>>
>>>>> I think Dora ought to get the new 3.10 kernel updates; this can go
>>>>> along side with it.
>>>>>
>>>>
>>>> Hi Otavio,
>>>>
>>>> I'm sorry, but I'm not sure whether we need update dora's kernel from
>>>> 3.10.11
>>>> to 3.10.17, since we seldom do the package update for a stable branch,
>>>> let's wait for others' comments:-).
>>>>
>>> This is correct, I do not think we will be updating the kernel for a stable
>>> branch.
>>
>> This is wrong. We should update to 3.10.17 as this is the LTS release,
>> otherwise what is the point of us using a LTS at all?
>
> Obviously I'll be maintaining the LTS release in master, but that
> maintenance gets a mix of bug fixes and features associated with the
> upstream maintenance + LTSI + BSP requirements.
>
> I've sent patches for the sustained branches in the past, and can
> do it here as well. But to do it properly, there really needs to be a
> dedicated kernel repository, since things like the LTSI import and
> new BSP support will be mixed into the tree if a single 3.10 repo
> is used.
>
> I'm willing to split the repo and do -stable korg release updates, but
> at some point there are only so many repositories and versions that
> should be maintained at one time. As to what that number is, I don't have
> a firm one in mind.
>
> At a minimum, it would be safe for the dora maintenance branch to take
> my latest kernel -stable imports, but in the not to distant future,
> it would need a separate maintenance stream.
>
> Cheers,
>
> Bruce
>
>>
>
>
>
Bruce Ashfield - Nov. 19, 2013, 5:18 a.m.
On Mon, Nov 18, 2013 at 10:34 PM, Robert Yang <liezhi.yang@windriver.com> wrote:
>
> Hi Bruce,
>
> Here is the log status of the kernel layer between dora and master (18
> commits).
>
> I wonder are there any serious problems that we must upgrade it ?
>
> $ git log dora..master --abbrev-commit --pretty=oneline meta/recipes-kernel
>
> 9a9a597 lttng-tools: Fixes incorrect path of ptest cases
> e1e1ee5 linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for
> Beagleboard
> 779dec9 linux-yocto-rt/3.10: fix ntp merge issue
> 89c986c linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for
> Beagleboard

You definitely want these. They fix USB and ethernet on the beagle

> 8a03d59 sysvinit: adjust boot sequence and remove hack from udev
> e9ec153 linux-yocto/3.10: fix qemuarm boot and spurious mips build warning
> baf005b linux-yocto/3.10: common-pc: add missing dependencies for BRCMSMAC
> dd361f0 linux-yocto/3.8: add crystalforest bsp legacy block drivers
> configurations
> 1fe7bca linux-yocto/3.10: haswell-wc and crystalforest support

Don't worry about the "add support" here, they are benign to all other goards.

> a23c7e9 linux-yocto-3.10: bump to 3.10.17 and -rt11

And this is where we need to be for the lttng fixes.

The other updates are also safe, you can't really pick and chose. Just take all
my SRCREV updates for 3.10 on master, since they are all bugfixes or minor
meta-data tweaks so far.

Bruce

> 3129619 lttng-tools: make ptest able to work on target
> 5e42796 recipes: Remove PR = r0 from all recipes
> 2570f13 perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for mips64
> 56cc2f2 linux-yocto/3.10: MinnowBoard support
> db1d37a perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for
> powerpc64
> cd414c0 kmod: Update to Rev 15 via git
> e83640d kmod: Add patch to fix seperate build dir of ptest
> 0f431a3 kmod-native: use bswap to work on older Linux hosts
>
> // Robert
>
>
> On 11/18/2013 11:59 PM, Bruce Ashfield wrote:
>>
>> On 13-11-18 10:50 AM, Otavio Salvador wrote:
>>>
>>> On Mon, Nov 18, 2013 at 1:31 PM, Saul Wold <saul.wold@intel.com> wrote:
>>>>
>>>> On 11/18/2013 06:50 AM, Robert Yang wrote:
>>>>>
>>>>>
>>>>>
>>>>> On 11/18/2013 10:18 PM, Otavio Salvador wrote:
>>>>>>
>>>>>>
>>>>>> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
>>>>>> <liezhi.yang@windriver.com> wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> This should be considered for Dora as well as it fixes the deadlock
>>>>>>>> and building with 3.12 kernels.
>>>>>>>>
>>>>>>>
>>>>>>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
>>>>>>> 3.10.11,
>>>>>>> and this is an update for lttng-modules, I'm not sure whether dora
>>>>>>> needs
>>>>>>> this.
>>>>>>>
>>>>>>> Add Saul in the CC list.
>>>>>>
>>>>>>
>>>>>>
>>>>>> I think Dora ought to get the new 3.10 kernel updates; this can go
>>>>>> along side with it.
>>>>>>
>>>>>
>>>>> Hi Otavio,
>>>>>
>>>>> I'm sorry, but I'm not sure whether we need update dora's kernel from
>>>>> 3.10.11
>>>>> to 3.10.17, since we seldom do the package update for a stable branch,
>>>>> let's wait for others' comments:-).
>>>>>
>>>> This is correct, I do not think we will be updating the kernel for a
>>>> stable
>>>> branch.
>>>
>>>
>>> This is wrong. We should update to 3.10.17 as this is the LTS release,
>>> otherwise what is the point of us using a LTS at all?
>>
>>
>> Obviously I'll be maintaining the LTS release in master, but that
>> maintenance gets a mix of bug fixes and features associated with the
>> upstream maintenance + LTSI + BSP requirements.
>>
>> I've sent patches for the sustained branches in the past, and can
>> do it here as well. But to do it properly, there really needs to be a
>> dedicated kernel repository, since things like the LTSI import and
>> new BSP support will be mixed into the tree if a single 3.10 repo
>> is used.
>>
>> I'm willing to split the repo and do -stable korg release updates, but
>> at some point there are only so many repositories and versions that
>> should be maintained at one time. As to what that number is, I don't have
>> a firm one in mind.
>>
>> At a minimum, it would be safe for the dora maintenance branch to take
>> my latest kernel -stable imports, but in the not to distant future,
>> it would need a separate maintenance stream.
>>
>> Cheers,
>>
>> Bruce
>>
>>>
>>
>>
>>
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
Robert Yang - Nov. 21, 2013, 3:05 a.m.
Hi Bruce,

Thanks, we are going to update the kernel from 3.10.11 to 3.10.17 for dora
if there is no objections.

Will we use another kernel source repo for dora ? Or just cherry-pick these
commits ?

// Robert

On 11/19/2013 01:18 PM, Bruce Ashfield wrote:
> On Mon, Nov 18, 2013 at 10:34 PM, Robert Yang <liezhi.yang@windriver.com> wrote:
>>
>> Hi Bruce,
>>
>> Here is the log status of the kernel layer between dora and master (18
>> commits).
>>
>> I wonder are there any serious problems that we must upgrade it ?
>>
>> $ git log dora..master --abbrev-commit --pretty=oneline meta/recipes-kernel
>>
>> 9a9a597 lttng-tools: Fixes incorrect path of ptest cases
>> e1e1ee5 linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for
>> Beagleboard
>> 779dec9 linux-yocto-rt/3.10: fix ntp merge issue
>> 89c986c linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for
>> Beagleboard
>
> You definitely want these. They fix USB and ethernet on the beagle
>
>> 8a03d59 sysvinit: adjust boot sequence and remove hack from udev
>> e9ec153 linux-yocto/3.10: fix qemuarm boot and spurious mips build warning
>> baf005b linux-yocto/3.10: common-pc: add missing dependencies for BRCMSMAC
>> dd361f0 linux-yocto/3.8: add crystalforest bsp legacy block drivers
>> configurations
>> 1fe7bca linux-yocto/3.10: haswell-wc and crystalforest support
>
> Don't worry about the "add support" here, they are benign to all other goards.
>
>> a23c7e9 linux-yocto-3.10: bump to 3.10.17 and -rt11
>
> And this is where we need to be for the lttng fixes.
>
> The other updates are also safe, you can't really pick and chose. Just take all
> my SRCREV updates for 3.10 on master, since they are all bugfixes or minor
> meta-data tweaks so far.
>
> Bruce
>
>> 3129619 lttng-tools: make ptest able to work on target
>> 5e42796 recipes: Remove PR = r0 from all recipes
>> 2570f13 perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for mips64
>> 56cc2f2 linux-yocto/3.10: MinnowBoard support
>> db1d37a perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for
>> powerpc64
>> cd414c0 kmod: Update to Rev 15 via git
>> e83640d kmod: Add patch to fix seperate build dir of ptest
>> 0f431a3 kmod-native: use bswap to work on older Linux hosts
>>
>> // Robert
>>
>>
>> On 11/18/2013 11:59 PM, Bruce Ashfield wrote:
>>>
>>> On 13-11-18 10:50 AM, Otavio Salvador wrote:
>>>>
>>>> On Mon, Nov 18, 2013 at 1:31 PM, Saul Wold <saul.wold@intel.com> wrote:
>>>>>
>>>>> On 11/18/2013 06:50 AM, Robert Yang wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>> On 11/18/2013 10:18 PM, Otavio Salvador wrote:
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
>>>>>>> <liezhi.yang@windriver.com> wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> This should be considered for Dora as well as it fixes the deadlock
>>>>>>>>> and building with 3.12 kernels.
>>>>>>>>>
>>>>>>>>
>>>>>>>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel is
>>>>>>>> 3.10.11,
>>>>>>>> and this is an update for lttng-modules, I'm not sure whether dora
>>>>>>>> needs
>>>>>>>> this.
>>>>>>>>
>>>>>>>> Add Saul in the CC list.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> I think Dora ought to get the new 3.10 kernel updates; this can go
>>>>>>> along side with it.
>>>>>>>
>>>>>>
>>>>>> Hi Otavio,
>>>>>>
>>>>>> I'm sorry, but I'm not sure whether we need update dora's kernel from
>>>>>> 3.10.11
>>>>>> to 3.10.17, since we seldom do the package update for a stable branch,
>>>>>> let's wait for others' comments:-).
>>>>>>
>>>>> This is correct, I do not think we will be updating the kernel for a
>>>>> stable
>>>>> branch.
>>>>
>>>>
>>>> This is wrong. We should update to 3.10.17 as this is the LTS release,
>>>> otherwise what is the point of us using a LTS at all?
>>>
>>>
>>> Obviously I'll be maintaining the LTS release in master, but that
>>> maintenance gets a mix of bug fixes and features associated with the
>>> upstream maintenance + LTSI + BSP requirements.
>>>
>>> I've sent patches for the sustained branches in the past, and can
>>> do it here as well. But to do it properly, there really needs to be a
>>> dedicated kernel repository, since things like the LTSI import and
>>> new BSP support will be mixed into the tree if a single 3.10 repo
>>> is used.
>>>
>>> I'm willing to split the repo and do -stable korg release updates, but
>>> at some point there are only so many repositories and versions that
>>> should be maintained at one time. As to what that number is, I don't have
>>> a firm one in mind.
>>>
>>> At a minimum, it would be safe for the dora maintenance branch to take
>>> my latest kernel -stable imports, but in the not to distant future,
>>> it would need a separate maintenance stream.
>>>
>>> Cheers,
>>>
>>> Bruce
>>>
>>>>
>>>
>>>
>>>
>> _______________________________________________
>> Openembedded-core mailing list
>> Openembedded-core@lists.openembedded.org
>> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>
>
>
Bruce Ashfield - Nov. 21, 2013, 6:05 a.m.
On Wed, Nov 20, 2013 at 10:05 PM, Robert Yang <liezhi.yang@windriver.com> wrote:
>
> Hi Bruce,
>
> Thanks, we are going to update the kernel from 3.10.11 to 3.10.17 for dora
> if there is no objections.
>
> Will we use another kernel source repo for dora ? Or just cherry-pick these
> commits ?

Just do the cherry picks. The repository is safe for the dora branch.
If that changes
in the future, we can fork it into a new repo.

Cheers,

Bruce

>
> // Robert
>
>
> On 11/19/2013 01:18 PM, Bruce Ashfield wrote:
>>
>> On Mon, Nov 18, 2013 at 10:34 PM, Robert Yang <liezhi.yang@windriver.com>
>> wrote:
>>>
>>>
>>> Hi Bruce,
>>>
>>> Here is the log status of the kernel layer between dora and master (18
>>> commits).
>>>
>>> I wonder are there any serious problems that we must upgrade it ?
>>>
>>> $ git log dora..master --abbrev-commit --pretty=oneline
>>> meta/recipes-kernel
>>>
>>> 9a9a597 lttng-tools: Fixes incorrect path of ptest cases
>>> e1e1ee5 linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for
>>> Beagleboard
>>> 779dec9 linux-yocto-rt/3.10: fix ntp merge issue
>>> 89c986c linux-yocto/3.10: meta: ARM: OMAP3: Add USB PHY driver for
>>> Beagleboard
>>
>>
>> You definitely want these. They fix USB and ethernet on the beagle
>>
>>> 8a03d59 sysvinit: adjust boot sequence and remove hack from udev
>>> e9ec153 linux-yocto/3.10: fix qemuarm boot and spurious mips build
>>> warning
>>> baf005b linux-yocto/3.10: common-pc: add missing dependencies for
>>> BRCMSMAC
>>> dd361f0 linux-yocto/3.8: add crystalforest bsp legacy block drivers
>>> configurations
>>> 1fe7bca linux-yocto/3.10: haswell-wc and crystalforest support
>>
>>
>> Don't worry about the "add support" here, they are benign to all other
>> goards.
>>
>>> a23c7e9 linux-yocto-3.10: bump to 3.10.17 and -rt11
>>
>>
>> And this is where we need to be for the lttng fixes.
>>
>> The other updates are also safe, you can't really pick and chose. Just
>> take all
>> my SRCREV updates for 3.10 on master, since they are all bugfixes or minor
>> meta-data tweaks so far.
>>
>> Bruce
>>
>>> 3129619 lttng-tools: make ptest able to work on target
>>> 5e42796 recipes: Remove PR = r0 from all recipes
>>> 2570f13 perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for
>>> mips64
>>> 56cc2f2 linux-yocto/3.10: MinnowBoard support
>>> db1d37a perf: flag __SANE_USERSPACE_TYPES__ to include int-ll64.h for
>>> powerpc64
>>> cd414c0 kmod: Update to Rev 15 via git
>>> e83640d kmod: Add patch to fix seperate build dir of ptest
>>> 0f431a3 kmod-native: use bswap to work on older Linux hosts
>>>
>>> // Robert
>>>
>>>
>>> On 11/18/2013 11:59 PM, Bruce Ashfield wrote:
>>>>
>>>>
>>>> On 13-11-18 10:50 AM, Otavio Salvador wrote:
>>>>>
>>>>>
>>>>> On Mon, Nov 18, 2013 at 1:31 PM, Saul Wold <saul.wold@intel.com> wrote:
>>>>>>
>>>>>>
>>>>>> On 11/18/2013 06:50 AM, Robert Yang wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On 11/18/2013 10:18 PM, Otavio Salvador wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, Nov 18, 2013 at 10:25 AM, Robert Yang
>>>>>>>> <liezhi.yang@windriver.com> wrote:
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> This should be considered for Dora as well as it fixes the
>>>>>>>>>> deadlock
>>>>>>>>>> and building with 3.12 kernels.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Hi Otavio, thanks for suggesting this for dora, but dora's kernel
>>>>>>>>> is
>>>>>>>>> 3.10.11,
>>>>>>>>> and this is an update for lttng-modules, I'm not sure whether dora
>>>>>>>>> needs
>>>>>>>>> this.
>>>>>>>>>
>>>>>>>>> Add Saul in the CC list.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> I think Dora ought to get the new 3.10 kernel updates; this can go
>>>>>>>> along side with it.
>>>>>>>>
>>>>>>>
>>>>>>> Hi Otavio,
>>>>>>>
>>>>>>> I'm sorry, but I'm not sure whether we need update dora's kernel from
>>>>>>> 3.10.11
>>>>>>> to 3.10.17, since we seldom do the package update for a stable
>>>>>>> branch,
>>>>>>> let's wait for others' comments:-).
>>>>>>>
>>>>>> This is correct, I do not think we will be updating the kernel for a
>>>>>> stable
>>>>>> branch.
>>>>>
>>>>>
>>>>>
>>>>> This is wrong. We should update to 3.10.17 as this is the LTS release,
>>>>> otherwise what is the point of us using a LTS at all?
>>>>
>>>>
>>>>
>>>> Obviously I'll be maintaining the LTS release in master, but that
>>>> maintenance gets a mix of bug fixes and features associated with the
>>>> upstream maintenance + LTSI + BSP requirements.
>>>>
>>>> I've sent patches for the sustained branches in the past, and can
>>>> do it here as well. But to do it properly, there really needs to be a
>>>> dedicated kernel repository, since things like the LTSI import and
>>>> new BSP support will be mixed into the tree if a single 3.10 repo
>>>> is used.
>>>>
>>>> I'm willing to split the repo and do -stable korg release updates, but
>>>> at some point there are only so many repositories and versions that
>>>> should be maintained at one time. As to what that number is, I don't
>>>> have
>>>> a firm one in mind.
>>>>
>>>> At a minimum, it would be safe for the dora maintenance branch to take
>>>> my latest kernel -stable imports, but in the not to distant future,
>>>> it would need a separate maintenance stream.
>>>>
>>>> Cheers,
>>>>
>>>> Bruce
>>>>
>>>>>
>>>>
>>>>
>>>>
>>> _______________________________________________
>>> Openembedded-core mailing list
>>> Openembedded-core@lists.openembedded.org
>>> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>>
>>
>>
>>
>

Patch

diff --git a/meta/recipes-kernel/lttng/lttng-modules_git.bb b/meta/recipes-kernel/lttng/lttng-modules_git.bb
index 789b5ec..e887a86 100644
--- a/meta/recipes-kernel/lttng/lttng-modules_git.bb
+++ b/meta/recipes-kernel/lttng/lttng-modules_git.bb
@@ -10,8 +10,8 @@  DEPENDS = "virtual/kernel"
 
 inherit module
 
-SRCREV = "78c8710bb8e6b7f6301d95afec0305f40fa063ad"
-PV = "2.3.0+git${SRCPV}"
+SRCREV = "18740750760fde894eac0cd4e3ef4d8c7c4ba8c3"
+PV = "2.3.2+git${SRCPV}"
 
 SRC_URI = "git://git.lttng.org/lttng-modules.git \
            file://lttng-modules-replace-KERNELDIR-with-KERNEL_SRC.patch"