* Compile error for boost_1.54.0 @ 2013-07-09 19:09 Chris Tapp 2013-07-09 20:19 ` Chris Tapp 0 siblings, 1 reply; 11+ messages in thread From: Chris Tapp @ 2013-07-09 19:09 UTC (permalink / raw) To: meta-freescale I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. Has anyone seen this before? Chris Tapp opensource@keylevel.com www.keylevel.com ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Compile error for boost_1.54.0 2013-07-09 19:09 Compile error for boost_1.54.0 Chris Tapp @ 2013-07-09 20:19 ` Chris Tapp 2013-07-09 20:21 ` Otavio Salvador 0 siblings, 1 reply; 11+ messages in thread From: Chris Tapp @ 2013-07-09 20:19 UTC (permalink / raw) To: meta-freescale On 9 Jul 2013, at 20:09, Chris Tapp wrote: > I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. > > The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. > > Has anyone seen this before? I also meant to say that adding: typedef unsigned long long uintptr_t; to atomic.hpp 'fixes' the build, but this is not a good solution ;-) Chris Tapp opensource@keylevel.com www.keylevel.com ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: Compile error for boost_1.54.0 2013-07-09 20:19 ` Chris Tapp @ 2013-07-09 20:21 ` Otavio Salvador 2013-07-09 20:25 ` [meta-freescale] " Chris Tapp 0 siblings, 1 reply; 11+ messages in thread From: Otavio Salvador @ 2013-07-09 20:21 UTC (permalink / raw) To: Chris Tapp; +Cc: meta-freescale@yoctoproject.org On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: > > On 9 Jul 2013, at 20:09, Chris Tapp wrote: > >> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >> >> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >> >> Has anyone seen this before? > > I also meant to say that adding: > > typedef unsigned long long uintptr_t; > > to atomic.hpp 'fixes' the build, but this is not a good solution ;-) It needs to be checked against normal Poky to ensure it is BSP specific; I doubt it is. -- Otavio Salvador O.S. Systems http://www.ossystems.com.br http://projetos.ossystems.com.br Mobile: +55 (53) 9981-7854 Mobile: +1 (347) 903-9750 ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [meta-freescale] Compile error for boost_1.54.0 2013-07-09 20:21 ` Otavio Salvador @ 2013-07-09 20:25 ` Chris Tapp 2013-07-09 20:50 ` Martin Jansa 0 siblings, 1 reply; 11+ messages in thread From: Chris Tapp @ 2013-07-09 20:25 UTC (permalink / raw) To: Otavio Salvador; +Cc: Yocto Discussion Mailing List Forwarding to Yocto mailing list: On 9 Jul 2013, at 21:21, Otavio Salvador wrote: > On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: >> >> On 9 Jul 2013, at 20:09, Chris Tapp wrote: >> >>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >>> >>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >>> >>> Has anyone seen this before? >> >> I also meant to say that adding: >> >> typedef unsigned long long uintptr_t; >> >> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) > > It needs to be checked against normal Poky to ensure it is BSP > specific; I doubt it is. > > -- > Otavio Salvador O.S. Systems > http://www.ossystems.com.br http://projetos.ossystems.com.br > Mobile: +55 (53) 9981-7854 Mobile: +1 (347) 903-9750 Chris Tapp opensource@keylevel.com www.keylevel.com ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [meta-freescale] Compile error for boost_1.54.0 2013-07-09 20:25 ` [meta-freescale] " Chris Tapp @ 2013-07-09 20:50 ` Martin Jansa 2013-07-10 3:17 ` Saul Wold 2013-07-10 5:04 ` Khem Raj 0 siblings, 2 replies; 11+ messages in thread From: Martin Jansa @ 2013-07-09 20:50 UTC (permalink / raw) To: Chris Tapp; +Cc: Yocto Discussion Mailing List, Otavio Salvador [-- Attachment #1: Type: text/plain, Size: 1141 bytes --] On Tue, Jul 09, 2013 at 09:25:30PM +0100, Chris Tapp wrote: > Forwarding to Yocto mailing list: > > On 9 Jul 2013, at 21:21, Otavio Salvador wrote: > > > On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: > >> > >> On 9 Jul 2013, at 20:09, Chris Tapp wrote: > >> > >>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. > >>> > >>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. > >>> > >>> Has anyone seen this before? > >> > >> I also meant to say that adding: > >> > >> typedef unsigned long long uintptr_t; > >> > >> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) > > > > It needs to be checked against normal Poky to ensure it is BSP > > specific; I doubt it is. See http://lists.openembedded.org/pipermail/openembedded-devel/2013-July/091331.html But maybe it wasn't caused by eglibc upgrade (and header cleanup in eglibc) but by boost upgrade. -- Martin 'JaMa' Jansa jabber: Martin.Jansa@gmail.com [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 205 bytes --] ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [meta-freescale] Compile error for boost_1.54.0 2013-07-09 20:50 ` Martin Jansa @ 2013-07-10 3:17 ` Saul Wold 2013-07-10 4:19 ` Khem Raj 2013-07-10 5:04 ` Khem Raj 1 sibling, 1 reply; 11+ messages in thread From: Saul Wold @ 2013-07-10 3:17 UTC (permalink / raw) To: Martin Jansa, Khem Raj; +Cc: Yocto Discussion Mailing List, Otavio Salvador On 07/09/2013 01:50 PM, Martin Jansa wrote: > On Tue, Jul 09, 2013 at 09:25:30PM +0100, Chris Tapp wrote: >> Forwarding to Yocto mailing list: >> >> On 9 Jul 2013, at 21:21, Otavio Salvador wrote: >> >>> On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: >>>> >>>> On 9 Jul 2013, at 20:09, Chris Tapp wrote: >>>> >>>>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >>>>> >>>>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >>>>> >>>>> Has anyone seen this before? >>>> >>>> I also meant to say that adding: >>>> >>>> typedef unsigned long long uintptr_t; >>>> >>>> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) >>> >>> It needs to be checked against normal Poky to ensure it is BSP >>> specific; I doubt it is. > > See > http://lists.openembedded.org/pipermail/openembedded-devel/2013-July/091331.html > > But maybe it wasn't caused by eglibc upgrade (and header cleanup in > eglibc) but by boost upgrade. > I did the boost update and with the older eglibc, which is why it seemed to work for me. I just did another build with the updated boost and both eglibc version 2.17 and 2.18, seems like the 2.18 version causes this breakage, as it built fine with 2.17, I also tried to revert the boost updated and build the older boost with 2.18 and it worked. So it must be something with the newer boost and eglibc in combination, more digging is required. Sau! > > > _______________________________________________ > yocto mailing list > yocto@yoctoproject.org > https://lists.yoctoproject.org/listinfo/yocto > ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [meta-freescale] Compile error for boost_1.54.0 2013-07-10 3:17 ` Saul Wold @ 2013-07-10 4:19 ` Khem Raj 0 siblings, 0 replies; 11+ messages in thread From: Khem Raj @ 2013-07-10 4:19 UTC (permalink / raw) To: Saul Wold; +Cc: Yocto Discussion Mailing List, Otavio Salvador On Jul 9, 2013, at 8:17 PM, Saul Wold <sgw@linux.intel.com> wrote: > On 07/09/2013 01:50 PM, Martin Jansa wrote: >> On Tue, Jul 09, 2013 at 09:25:30PM +0100, Chris Tapp wrote: >>> Forwarding to Yocto mailing list: >>> >>> On 9 Jul 2013, at 21:21, Otavio Salvador wrote: >>> >>>> On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: >>>>> >>>>> On 9 Jul 2013, at 20:09, Chris Tapp wrote: >>>>> >>>>>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >>>>>> >>>>>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >>>>>> >>>>>> Has anyone seen this before? >>>>> >>>>> I also meant to say that adding: >>>>> >>>>> typedef unsigned long long uintptr_t; >>>>> >>>>> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) >>>> >>>> It needs to be checked against normal Poky to ensure it is BSP >>>> specific; I doubt it is. >> >> See >> http://lists.openembedded.org/pipermail/openembedded-devel/2013-July/091331.html >> >> But maybe it wasn't caused by eglibc upgrade (and header cleanup in >> eglibc) but by boost upgrade. >> > I did the boost update and with the older eglibc, which is why it seemed to work for me. > > I just did another build with the updated boost and both eglibc version 2.17 and 2.18, seems like the 2.18 version causes this breakage, as it built fine with 2.17, I also tried to revert the boost updated and build the older boost with 2.18 and it worked. > > So it must be something with the newer boost and eglibc in combination, more digging is required. > its related to https://bugzilla.yoctoproject.org/show_bug.cgi?id=4812 > Sau! > >> >> >> _______________________________________________ >> yocto mailing list >> yocto@yoctoproject.org >> https://lists.yoctoproject.org/listinfo/yocto >> ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [meta-freescale] Compile error for boost_1.54.0 2013-07-09 20:50 ` Martin Jansa 2013-07-10 3:17 ` Saul Wold @ 2013-07-10 5:04 ` Khem Raj 2013-07-11 19:27 ` Saul Wold 1 sibling, 1 reply; 11+ messages in thread From: Khem Raj @ 2013-07-10 5:04 UTC (permalink / raw) To: Martin Jansa; +Cc: Yocto Discussion Mailing List, Otavio Salvador On Jul 9, 2013, at 1:50 PM, Martin Jansa <martin.jansa@gmail.com> wrote: > On Tue, Jul 09, 2013 at 09:25:30PM +0100, Chris Tapp wrote: >> Forwarding to Yocto mailing list: >> >> On 9 Jul 2013, at 21:21, Otavio Salvador wrote: >> >>> On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: >>>> >>>> On 9 Jul 2013, at 20:09, Chris Tapp wrote: >>>> >>>>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >>>>> >>>>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >>>>> >>>>> Has anyone seen this before? >>>> >>>> I also meant to say that adding: >>>> >>>> typedef unsigned long long uintptr_t; >>>> >>>> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) >>> >>> It needs to be checked against normal Poky to ensure it is BSP >>> specific; I doubt it is. > > See > http://lists.openembedded.org/pipermail/openembedded-devel/2013-July/091331.html in above thread answering to your question if you want to include <cstblib> you need to specify -std=gnu++0x or -std=c++0x to cppflags. above container is not supported prior to c++0x standard > > But maybe it wasn't caused by eglibc upgrade (and header cleanup in > eglibc) but by boost upgrade. > > -- > Martin 'JaMa' Jansa jabber: Martin.Jansa@gmail.com > _______________________________________________ > yocto mailing list > yocto@yoctoproject.org > https://lists.yoctoproject.org/listinfo/yocto ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [meta-freescale] Compile error for boost_1.54.0 2013-07-10 5:04 ` Khem Raj @ 2013-07-11 19:27 ` Saul Wold 2013-07-11 19:38 ` [yocto] " Chris Tapp 0 siblings, 1 reply; 11+ messages in thread From: Saul Wold @ 2013-07-11 19:27 UTC (permalink / raw) To: Khem Raj; +Cc: Yocto Discussion Mailing List, Otavio Salvador OK, I finally tracked this bugger down to the eglibc update to 2.18 and specifically > ../../../eglibc/2.18-r0/eglibc-2.18/libc/ChangeLog: * include/features.h (__GLIBC_HAVE_LONG_LONG): Remove. Turns out that in boost code there is a check for the __GLIBC_HAVE_LONG_LONG in cstdint.hpp and this was not getting triggered any more. The include issue for g++ was kind of a red herring in this case, that issue still exists, but this is a different problem. And there is a patch for that in the SVN master, I will pull it! Sau! On 07/09/2013 10:04 PM, Khem Raj wrote: > > On Jul 9, 2013, at 1:50 PM, Martin Jansa <martin.jansa@gmail.com> wrote: > >> On Tue, Jul 09, 2013 at 09:25:30PM +0100, Chris Tapp wrote: >>> Forwarding to Yocto mailing list: >>> >>> On 9 Jul 2013, at 21:21, Otavio Salvador wrote: >>> >>>> On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: >>>>> >>>>> On 9 Jul 2013, at 20:09, Chris Tapp wrote: >>>>> >>>>>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >>>>>> >>>>>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >>>>>> >>>>>> Has anyone seen this before? >>>>> >>>>> I also meant to say that adding: >>>>> >>>>> typedef unsigned long long uintptr_t; >>>>> >>>>> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) >>>> >>>> It needs to be checked against normal Poky to ensure it is BSP >>>> specific; I doubt it is. >> >> See >> http://lists.openembedded.org/pipermail/openembedded-devel/2013-July/091331.html > > in above thread answering to your question if you want to include <cstblib> you need to specify > -std=gnu++0x or -std=c++0x to cppflags. above container is not supported prior to c++0x standard > > >> >> But maybe it wasn't caused by eglibc upgrade (and header cleanup in >> eglibc) but by boost upgrade. >> >> -- >> Martin 'JaMa' Jansa jabber: Martin.Jansa@gmail.com >> _______________________________________________ >> yocto mailing list >> yocto@yoctoproject.org >> https://lists.yoctoproject.org/listinfo/yocto > > _______________________________________________ > yocto mailing list > yocto@yoctoproject.org > https://lists.yoctoproject.org/listinfo/yocto > > ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [meta-freescale] Compile error for boost_1.54.0 2013-07-11 19:27 ` Saul Wold @ 2013-07-11 19:38 ` Chris Tapp 0 siblings, 0 replies; 11+ messages in thread From: Chris Tapp @ 2013-07-11 19:38 UTC (permalink / raw) To: Saul Wold; +Cc: meta-freescale, Yocto Discussion Mailing List, Otavio Salvador On 11 Jul 2013, at 20:27, Saul Wold wrote: > > OK, I finally tracked this bugger down to the eglibc update to 2.18 and specifically > >> ../../../eglibc/2.18-r0/eglibc-2.18/libc/ChangeLog: * include/features.h (__GLIBC_HAVE_LONG_LONG): Remove. > > Turns out that in boost code there is a check for the __GLIBC_HAVE_LONG_LONG in cstdint.hpp and this was not getting triggered any more. > > The include issue for g++ was kind of a red herring in this case, that issue still exists, but this is a different problem. And there is a patch for that in the SVN master, I will pull it! Thanks Saul :-) > Sau! > > > On 07/09/2013 10:04 PM, Khem Raj wrote: >> >> On Jul 9, 2013, at 1:50 PM, Martin Jansa <martin.jansa@gmail.com> wrote: >> >>> On Tue, Jul 09, 2013 at 09:25:30PM +0100, Chris Tapp wrote: >>>> Forwarding to Yocto mailing list: >>>> >>>> On 9 Jul 2013, at 21:21, Otavio Salvador wrote: >>>> >>>>> On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: >>>>>> >>>>>> On 9 Jul 2013, at 20:09, Chris Tapp wrote: >>>>>> >>>>>>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >>>>>>> >>>>>>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >>>>>>> >>>>>>> Has anyone seen this before? >>>>>> >>>>>> I also meant to say that adding: >>>>>> >>>>>> typedef unsigned long long uintptr_t; >>>>>> >>>>>> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) >>>>> >>>>> It needs to be checked against normal Poky to ensure it is BSP >>>>> specific; I doubt it is. >>> >>> See >>> http://lists.openembedded.org/pipermail/openembedded-devel/2013-July/091331.html >> >> in above thread answering to your question if you want to include <cstblib> you need to specify >> -std=gnu++0x or -std=c++0x to cppflags. above container is not supported prior to c++0x standard >> >> >>> >>> But maybe it wasn't caused by eglibc upgrade (and header cleanup in >>> eglibc) but by boost upgrade. >>> >>> -- >>> Martin 'JaMa' Jansa jabber: Martin.Jansa@gmail.com >>> _______________________________________________ >>> yocto mailing list >>> yocto@yoctoproject.org >>> https://lists.yoctoproject.org/listinfo/yocto >> >> _______________________________________________ >> yocto mailing list >> yocto@yoctoproject.org >> https://lists.yoctoproject.org/listinfo/yocto >> >> > _______________________________________________ > yocto mailing list > yocto@yoctoproject.org > https://lists.yoctoproject.org/listinfo/yocto Chris Tapp opensource@keylevel.com www.keylevel.com ^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [yocto] Compile error for boost_1.54.0 @ 2013-07-11 19:38 ` Chris Tapp 0 siblings, 0 replies; 11+ messages in thread From: Chris Tapp @ 2013-07-11 19:38 UTC (permalink / raw) To: Saul Wold Cc: meta-freescale, Yocto Discussion Mailing List, Khem Raj, Otavio Salvador On 11 Jul 2013, at 20:27, Saul Wold wrote: > > OK, I finally tracked this bugger down to the eglibc update to 2.18 and specifically > >> ../../../eglibc/2.18-r0/eglibc-2.18/libc/ChangeLog: * include/features.h (__GLIBC_HAVE_LONG_LONG): Remove. > > Turns out that in boost code there is a check for the __GLIBC_HAVE_LONG_LONG in cstdint.hpp and this was not getting triggered any more. > > The include issue for g++ was kind of a red herring in this case, that issue still exists, but this is a different problem. And there is a patch for that in the SVN master, I will pull it! Thanks Saul :-) > Sau! > > > On 07/09/2013 10:04 PM, Khem Raj wrote: >> >> On Jul 9, 2013, at 1:50 PM, Martin Jansa <martin.jansa@gmail.com> wrote: >> >>> On Tue, Jul 09, 2013 at 09:25:30PM +0100, Chris Tapp wrote: >>>> Forwarding to Yocto mailing list: >>>> >>>> On 9 Jul 2013, at 21:21, Otavio Salvador wrote: >>>> >>>>> On Tue, Jul 9, 2013 at 5:19 PM, Chris Tapp <opensource@keylevel.com> wrote: >>>>>> >>>>>> On 9 Jul 2013, at 20:09, Chris Tapp wrote: >>>>>> >>>>>>> I'm getting a failure in do_compile when building boost_1.54.0 using master-next for the wandboard-quad. >>>>>>> >>>>>>> The log is about 1.5MB, but the important bit seems to be that uintptr_t isn't defined but is required by /boost/atomic/atomic.hpp. >>>>>>> >>>>>>> Has anyone seen this before? >>>>>> >>>>>> I also meant to say that adding: >>>>>> >>>>>> typedef unsigned long long uintptr_t; >>>>>> >>>>>> to atomic.hpp 'fixes' the build, but this is not a good solution ;-) >>>>> >>>>> It needs to be checked against normal Poky to ensure it is BSP >>>>> specific; I doubt it is. >>> >>> See >>> http://lists.openembedded.org/pipermail/openembedded-devel/2013-July/091331.html >> >> in above thread answering to your question if you want to include <cstblib> you need to specify >> -std=gnu++0x or -std=c++0x to cppflags. above container is not supported prior to c++0x standard >> >> >>> >>> But maybe it wasn't caused by eglibc upgrade (and header cleanup in >>> eglibc) but by boost upgrade. >>> >>> -- >>> Martin 'JaMa' Jansa jabber: Martin.Jansa@gmail.com >>> _______________________________________________ >>> yocto mailing list >>> yocto@yoctoproject.org >>> https://lists.yoctoproject.org/listinfo/yocto >> >> _______________________________________________ >> yocto mailing list >> yocto@yoctoproject.org >> https://lists.yoctoproject.org/listinfo/yocto >> >> > _______________________________________________ > yocto mailing list > yocto@yoctoproject.org > https://lists.yoctoproject.org/listinfo/yocto Chris Tapp opensource@keylevel.com www.keylevel.com ^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2013-07-11 19:38 UTC | newest] Thread overview: 11+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2013-07-09 19:09 Compile error for boost_1.54.0 Chris Tapp 2013-07-09 20:19 ` Chris Tapp 2013-07-09 20:21 ` Otavio Salvador 2013-07-09 20:25 ` [meta-freescale] " Chris Tapp 2013-07-09 20:50 ` Martin Jansa 2013-07-10 3:17 ` Saul Wold 2013-07-10 4:19 ` Khem Raj 2013-07-10 5:04 ` Khem Raj 2013-07-11 19:27 ` Saul Wold 2013-07-11 19:38 ` Chris Tapp 2013-07-11 19:38 ` [yocto] " Chris Tapp
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.