[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Thomas Petazzoni]

Hello Philipp,

Thanks for taking a look at your autobuilder failures! Please try to
Cc: the Buildroot mailing list on such discussions.

On Mon, 3 Dec 2018 12:26:33 +0100, Philipp Richter wrote:
> In file included from peer_connection.cpp:56:0:
> ../include/libtorrent/peer_connection.hpp:550:8: error: generic thunk
> code fails for method ?virtual void
> libtorrent::peer_connection::_ZThn96_NK10libtorrent15peer_connection8peer_logENS_14peer_log_alert11direction_tEPKcS4_z(libtorrent::peer_log_alert::direction_t,
> const char*, const char*, ...) const? which uses ?...?
>    void peer_log(peer_log_alert::direction_t direction
>         ^
> 
> So there's an issue with the variadic function maybe because it's
> overriding a virtual variadic function ?
> According to this https://bugzilla.redhat.com/show_bug.cgi?id=1511021
> adding "[[gnu::optimize("O0")]]" to the function might be a
> workaround.
> Or maybe a newer gcc though for openrisc it seems to be stuck with gcc 5.
> I'll check it out and report back.

Note that it also happened on Xtensa:
http://autobuild.buildroot.net/results/38a/38a5481dff972954c02a74ab19f37e71d5f6510e/build-end.log,
which uses gcc 7.x.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Bartosz Golaszewski]

pon., 3 gru 2018 o 13:45 Thomas Petazzoni
<thomas.petazzoni@bootlin.com> napisa?(a):
>
> Hello Philipp,
>
> Thanks for taking a look at your autobuilder failures! Please try to
> Cc: the Buildroot mailing list on such discussions.
>
> On Mon, 3 Dec 2018 12:26:33 +0100, Philipp Richter wrote:
> > In file included from peer_connection.cpp:56:0:
> > ../include/libtorrent/peer_connection.hpp:550:8: error: generic thunk
> > code fails for method ?virtual void
> > libtorrent::peer_connection::_ZThn96_NK10libtorrent15peer_connection8peer_logENS_14peer_log_alert11direction_tEPKcS4_z(libtorrent::peer_log_alert::direction_t,
> > const char*, const char*, ...) const? which uses ?...?
> >    void peer_log(peer_log_alert::direction_t direction
> >         ^
> >
> > So there's an issue with the variadic function maybe because it's
> > overriding a virtual variadic function ?
> > According to this https://bugzilla.redhat.com/show_bug.cgi?id=1511021
> > adding "[[gnu::optimize("O0")]]" to the function might be a
> > workaround.
> > Or maybe a newer gcc though for openrisc it seems to be stuck with gcc 5.
> > I'll check it out and report back.
>
> Note that it also happened on Xtensa:
> http://autobuild.buildroot.net/results/38a/38a5481dff972954c02a74ab19f37e71d5f6510e/build-end.log,
> which uses gcc 7.x.
>
> Best regards,
>
> Thomas
> --
> Thomas Petazzoni, CTO, Bootlin
> Embedded Linux and Kernel engineering
> https://bootlin.com
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot

Hi Thomas,

this failure (and others in pulseview) are due to libsigrokdecode.h
being included when libsigrokdecode is not built. It's a bug in
pulseview. I already have a patch and will submit it later today both
for buildroot as well as upstream.

Best regards,
Bartosz Golaszewski

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Thomas Petazzoni]

Hello Bartosz,

On Mon, 3 Dec 2018 14:37:05 +0100, Bartosz Golaszewski wrote:

> this failure (and others in pulseview) are due to libsigrokdecode.h
> being included when libsigrokdecode is not built. It's a bug in
> pulseview. I already have a patch and will submit it later today both
> for buildroot as well as upstream.

Well, the error that Philipp points to is in libtorrent-rasterbar,
which has nothing to do with libsigrokdecode, and quite probably
doesn't use it. Are you sure the issue is related ?

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Bartosz Golaszewski]

pon., 3 gru 2018 o 14:45 Thomas Petazzoni
<thomas.petazzoni@bootlin.com> napisa?(a):
>
> Hello Bartosz,
>
> On Mon, 3 Dec 2018 14:37:05 +0100, Bartosz Golaszewski wrote:
>
> > this failure (and others in pulseview) are due to libsigrokdecode.h
> > being included when libsigrokdecode is not built. It's a bug in
> > pulseview. I already have a patch and will submit it later today both
> > for buildroot as well as upstream.
>
> Well, the error that Philipp points to is in libtorrent-rasterbar,
> which has nothing to do with libsigrokdecode, and quite probably
> doesn't use it. Are you sure the issue is related ?
>

What I meant was that I found the issue and it has nothing to do with
libtorrent. I should have expressed myself clearer.

Bart

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Thomas Petazzoni]

Hello,

On Mon, 3 Dec 2018 14:51:05 +0100, Bartosz Golaszewski wrote:

> > > this failure (and others in pulseview) are due to libsigrokdecode.h
> > > being included when libsigrokdecode is not built. It's a bug in
> > > pulseview. I already have a patch and will submit it later today both
> > > for buildroot as well as upstream.  
> >
> > Well, the error that Philipp points to is in libtorrent-rasterbar,
> > which has nothing to do with libsigrokdecode, and quite probably
> > doesn't use it. Are you sure the issue is related ?
> 
> What I meant was that I found the issue and it has nothing to do with
> libtorrent. I should have expressed myself clearer.

How can a bug in pulseview/libsigrokdecode after libttorrent-rasterbar ?

See the build failures we are talking about, they affect the
libtorrent-rasterbar package:

  http://autobuild.buildroot.net/?reason=libtorrent-rasterbar%

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Bartosz Golaszewski]

pon., 3 gru 2018 o 14:56 Thomas Petazzoni
<thomas.petazzoni@bootlin.com> napisa?(a):
>
> Hello,
>
> On Mon, 3 Dec 2018 14:51:05 +0100, Bartosz Golaszewski wrote:
>
> > > > this failure (and others in pulseview) are due to libsigrokdecode.h
> > > > being included when libsigrokdecode is not built. It's a bug in
> > > > pulseview. I already have a patch and will submit it later today both
> > > > for buildroot as well as upstream.
> > >
> > > Well, the error that Philipp points to is in libtorrent-rasterbar,
> > > which has nothing to do with libsigrokdecode, and quite probably
> > > doesn't use it. Are you sure the issue is related ?
> >
> > What I meant was that I found the issue and it has nothing to do with
> > libtorrent. I should have expressed myself clearer.
>
> How can a bug in pulseview/libsigrokdecode after libttorrent-rasterbar ?
>
> See the build failures we are talking about, they affect the
> libtorrent-rasterbar package:
>

I'm not following you here I'm afraid. This has nothing to do with any
other package - this is purely a pulseview bug. It appears after my
previous version bump and is confined to pulseview.

Bart

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Thomas Petazzoni]

Hello,

On Mon, 3 Dec 2018 15:25:25 +0100, Bartosz Golaszewski wrote:

> > How can a bug in pulseview/libsigrokdecode after libttorrent-rasterbar ?
> >
> > See the build failures we are talking about, they affect the
> > libtorrent-rasterbar package:
> >  
> 
> I'm not following you here I'm afraid. This has nothing to do with any
> other package - this is purely a pulseview bug. It appears after my
> previous version bump and is confined to pulseview.

Are you looking at:

  http://autobuild.buildroot.net/results/753/753059dd52da9ec2aa1b06f3b64b0ab5ee3e059a/build-end.log

Let me paste the relevant part:

In file included from peer_connection.cpp:56:0:
../include/libtorrent/peer_connection.hpp:550:8: error: generic thunk code fails for method 'virtual void libtorrent::peer_connection::_ZThn96_NK10libtorrent15peer_connection8peer_logENS_14peer_log_alert11direction_tEPKcS4_z(libtorrent::peer_log_alert::direction_t, const char*, const char*, ...) const' which uses '...'
   void peer_log(peer_log_alert::direction_t direction
        ^~~~~~~~
make[2]: *** [peer_connection.lo] Error 1
make[2]: *** Waiting for unfinished jobs....
make[2]: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11'
make: *** [/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11/.stamp_built] Error 2
make: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/buildroot'

Where do you see a pulseview issue ?

Thomas
-- 
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Bartosz Golaszewski]

pon., 3 gru 2018 o 15:34 Thomas Petazzoni
<thomas.petazzoni@bootlin.com> napisa?(a):
>
> Hello,
>
> On Mon, 3 Dec 2018 15:25:25 +0100, Bartosz Golaszewski wrote:
>
> > > How can a bug in pulseview/libsigrokdecode after libttorrent-rasterbar ?
> > >
> > > See the build failures we are talking about, they affect the
> > > libtorrent-rasterbar package:
> > >
> >
> > I'm not following you here I'm afraid. This has nothing to do with any
> > other package - this is purely a pulseview bug. It appears after my
> > previous version bump and is confined to pulseview.
>
> Are you looking at:
>
>   http://autobuild.buildroot.net/results/753/753059dd52da9ec2aa1b06f3b64b0ab5ee3e059a/build-end.log
>
> Let me paste the relevant part:
>
> In file included from peer_connection.cpp:56:0:
> ../include/libtorrent/peer_connection.hpp:550:8: error: generic thunk code fails for method 'virtual void libtorrent::peer_connection::_ZThn96_NK10libtorrent15peer_connection8peer_logENS_14peer_log_alert11direction_tEPKcS4_z(libtorrent::peer_log_alert::direction_t, const char*, const char*, ...) const' which uses '...'
>    void peer_log(peer_log_alert::direction_t direction
>         ^~~~~~~~
> make[2]: *** [peer_connection.lo] Error 1
> make[2]: *** Waiting for unfinished jobs....
> make[2]: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11/src'
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11'
> make: *** [/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11/.stamp_built] Error 2
> make: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/buildroot'
>
> Where do you see a pulseview issue ?
>

Eek I was looking at
http://autobuild.buildroot.net/results/79ed29bdd1eec9bf7cb5003202da5ceaeb8101b6/
why am I responding here? I have no idea. :)

Anyway, patch for pulseview sent. Sorry for the confusion.

Bart

[Buildroot] [autobuild.buildroot.net] Your build results for 2018-12-02

[Philipp Richter]

On Mon, 3 Dec 2018 at 13:45, Thomas Petazzoni
<thomas.petazzoni@bootlin.com> wrote:
>
> Hello Philipp,
>
> Thanks for taking a look at your autobuilder failures! Please try to
> Cc: the Buildroot mailing list on such discussions.

Right, I'll remember that, thank you again.

On these platforms it seems like gcc is missing something to support
this ? I'm not sure how to fix this.
There's the --disable-logging configuration flag that could be forced
on those platforms I guess, not very subtle but compilation goes
through.

Regards,
Philipp Richter

On Mon, 3 Dec 2018 at 15:37, Bartosz Golaszewski <brgl@bgdev.pl> wrote:
>
> pon., 3 gru 2018 o 15:34 Thomas Petazzoni
> <thomas.petazzoni@bootlin.com> napisa?(a):
> >
> > Hello,
> >
> > On Mon, 3 Dec 2018 15:25:25 +0100, Bartosz Golaszewski wrote:
> >
> > > > How can a bug in pulseview/libsigrokdecode after libttorrent-rasterbar ?
> > > >
> > > > See the build failures we are talking about, they affect the
> > > > libtorrent-rasterbar package:
> > > >
> > >
> > > I'm not following you here I'm afraid. This has nothing to do with any
> > > other package - this is purely a pulseview bug. It appears after my
> > > previous version bump and is confined to pulseview.
> >
> > Are you looking at:
> >
> >   http://autobuild.buildroot.net/results/753/753059dd52da9ec2aa1b06f3b64b0ab5ee3e059a/build-end.log
> >
> > Let me paste the relevant part:
> >
> > In file included from peer_connection.cpp:56:0:
> > ../include/libtorrent/peer_connection.hpp:550:8: error: generic thunk code fails for method 'virtual void libtorrent::peer_connection::_ZThn96_NK10libtorrent15peer_connection8peer_logENS_14peer_log_alert11direction_tEPKcS4_z(libtorrent::peer_log_alert::direction_t, const char*, const char*, ...) const' which uses '...'
> >    void peer_log(peer_log_alert::direction_t direction
> >         ^~~~~~~~
> > make[2]: *** [peer_connection.lo] Error 1
> > make[2]: *** Waiting for unfinished jobs....
> > make[2]: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11/src'
> > make[1]: *** [all-recursive] Error 1
> > make[1]: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11'
> > make: *** [/home/dawncrow/buildroot-test/scripts/instance-0/output/build/libtorrent-rasterbar-1.1.11/.stamp_built] Error 2
> > make: Leaving directory `/home/dawncrow/buildroot-test/scripts/instance-0/buildroot'
> >
> > Where do you see a pulseview issue ?
> >
>
> Eek I was looking at
> http://autobuild.buildroot.net/results/79ed29bdd1eec9bf7cb5003202da5ceaeb8101b6/
> why am I responding here? I have no idea. :)
>
> Anyway, patch for pulseview sent. Sorry for the confusion.
>
> Bart