Linphone

Hello Guix!

I am creating this thread for Linphone packaging. I will be sending patches one by one with-in this thread. :-)

Regards,

RG.

Missing definition of libantlr3c?

block 40264 by 40307

quit

block 40264 by 40285

block 40264 by 40291

block 40264 by 40348

block 40264 by 40349

:-)

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

This has now been reviewed & merged.

Hello Raghav,

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

^ You can drop the 'packages: linphone:' part

and add a dot.

Please only activate the switches which need to be activated (let's not

repeat the default values to keep the noise level/verbosity as low as

possible). You can usually see the defaults by inspecting the

CMakeLists.txt file at the root of the project.

^ a

The first 3 commits of this series should be squashed together.

Also, please run the indent-code.el script for proper indentation.

Thank you!

Maxim

Hi Raghav,

belle-sip is not reproducible.

$ guix build --rounds=2 belle-sip

[...]

phase `compress-documentation' succeeded after 0.0 seconds

output ‘/gnu/store/rykd8wp9ibjf3ad4nlg15vmyzgq9izsp-belle-sip-1.6.3’ of ‘/gnu/store/lqdmaxvlhxk25jr7whk5bvi0y1zbzp5a-belle-sip-1.6.3.drv’ differs from ‘/gnu/store/rykd8wp9ibjf3ad4nlg15vmyzgq9izsp-belle-sip-1.6.3-check’ from previous round

build of /gnu/store/lqdmaxvlhxk25jr7whk5bvi0y1zbzp5a-belle-sip-1.6.3.drv failed

$ diffoscope --exclude-directory-metadata /gnu/store/rykd8wp9ibjf3ad4nlg15vmyzgq9izsp-belle-sip-1.6.3 /gnu/store/rykd8wp9ibjf3ad4nlg15vmyzgq9izsp-belle-sip-1.6.3-check

[...]

I'm not saying that this is a blocker for merging, but basically libbellesip.so (and libbellesip.so.0) doesn't always come out the same. Spooky.

Hello Maxim and Danny!

At this point for linphoneqt a.k.a linphone-desktop, I am facing following issues.

When I build *without* `-DENABLE_DBUS=YES`and run the program, I get:

QSocketNotifier: Socket notifiers cannot be enabled or disabled from another thread

QMutex: destroying locked mutex

When I build *with* `-DENABLE_DBUS=YES` and run the program, I get:

Segmentation Fault (Core Dumped)

I think the following patch is relevant, but when I use it, doesn't get successfully patched during the build.

Any ideas on how to fix this situation?

Thank you!

Regards,

RG.

Hi Danny!

Nice catch!

Yeah, that is really is spooky. I even disabled the tests as it required network access. Is it possible that there are multiple values set for any environment variable, and it uses different values randomly during each build?

Regards,

RG.

Hi Danny!

I suspect it could be due to having two different definitions for bctoolbox and ortp. Could you try belle-sip after applying the patch http://debbugs.gnu.org/cgi/bugreport.cgi?bug=40326

Regards,

RG.

Hello Maxim and Danny!

So I fixed "QSocketNotifier" error in version 4 patch (14-add-linphoneqt-v4). I still don't know how to fix "QMutex" error.

Regards,

RG.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit d5a8e391ac2546241d0ff60c9a1fd2fb8ff57536.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit c8f69ec27b4d492f63edc71f0a83ec413ea85b96.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit bfaabfebac0fc6c3fba2ed4358556b393ef535ce.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit 63c73bcad0ccd44363611162010b24e2b60629f4.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit 28f6f1e0da2481ce854a893d72d6bdfc5ce97952.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit 02d5d72cfc525e1ec53a02f3a201872f261c7c37.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit 13f401d3665a246f5e7d3a9076fdedfe744dd3fd.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit f3697439cd226f769cdc5ce962f08ad3a1a8bfd6.

Hi Raghav,

On Thu, 02 Apr 2020 17:19:29 +0000

"Raghav Gururajan" <raghavgururajan@disroot.org> wrote:

I tried and it didn't help.

However, I found and fixed the problem:

From 1f847861de4ab73011d6243fb652f098e16106cb Mon Sep 17 00:00:00 2001

Tags: patch

---

gnu/packages/linphone.scm | 42 +++++++++++++++++++++++++++++++++++++++

1 file changed, 42 insertions(+)

Hello everyone,

On Thu, 02 Apr 2020 00:20:04 +0000

"Raghav Gururajan" <raghavgururajan@disroot.org> wrote:

Yes, but why? What would happen if we didn't do it?

I'm not sure what the outcome of the latest discussion on carrying patches was.

Should we carry patches in Guix now or fetch them from external source?

Uhhhhhhhhh if that's a library, it should have been compiled with "-fPIC", no?

Which library would need recompilation with "-fPIC"?

Hi Danny!

Otherwise, we get `-Werror` flag related compilation errors.

I think it's good as long as we use hash to verify. :-)

I am not sure about this one. Since it is a feature, I was planning on to deal this later and disable gsm for now. I hope it is okay.

Regards,

RG.

Hello Maxim and Danny!

I have now fixed the "Qmutex error" in version 6 patch (14-add-linphoneqt-v6).

Regards,

RG.

Hello Danny and Raghav,

Danny Milosavljevic <dannym@scratchpost.org> writes:

Nice find! Although note that this doesn't exist in the latest code,

and that the latest release is nearly 3 years old (2017-07-21). Perhaps

we should just try to package the latest master? Their git history

suggests many bug fixes, including a memory leak.

Maxim

In their mail list, one of the dev mentioned that there is going to be a release this year. Also, for consistency, I think it is better that we stick to release tarballs and apply critical patches from master. :-)

Regards,

RG.

Hello Raghav,

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

Could you please clarify that in a comment above the patch?

The problem is that metadata or format of auto-generated patches may

change on a whim. In past discussions it was decided best to not use

them (that is, it's best to embed a copy of it in Guix sources) [0].

[0] https://lists.gnu.org/archive/html/guix-devel/2018-11/msg00554.html

I'll post more comments in your original message.

Maxim

Hi Raghav!

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

The variable name and package name should be "mediastreamer2", as per

upstream.

You can just use the "ENABLE_STRICT=NO", which will disable -Werror and

make the package more resilient in the face of future compiler upgrades

in Guix. No more need for that patch :-).

Usually we try to offer the most features possible. I don't see why we

shouldn't add SRTP support since we should already have the required

dependencies.

Also, I see there's a ENABLE_UNIT_TESTS=YES option for unit tests, which

we should enable and try to run.

Not a merge breaker, but it'd make sense to have GSM for telephony

focused software :-).

There are some strange characters in there (around 'powerful'). Please

edit them out.

Maxim

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

Sounds reasonable! Thanks for the information.

Maxim

block 40264 by 40445

Hi Maxim!

I fixed everything in 11-add-mediastreamer-v3. :-)

Regards,

RG.

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

Is the mediastreamer-srtp2 patch really required? I could build the

package successfully (with libsrtp) without it. Please rename the

patches mediastream2 to reflect the updated package name (the '2'

suffix).

Also, patches must be registered in gnu/local.mk (in the dist_patch_DATA

variable), otherwise they won't be included in the sources of Guix.

Could you investigate the ENABLE_UNIT_TESTS=yes option?

I still see strange characters here: ^ ^

Otherwise, LGTM.

Maxim

Does

ENABLE_STRICT=NO

instead of mediastreamer-msfactory.patch work or not?

Pushed this patch to guix master as commit e7859da11371e9bd0fffc5235b9e980cccf4b239.

Hi,

On Sun, 5 Apr 2020 20:29:40 +0200

Danny Milosavljevic <dannym@scratchpost.org> wrote:

Yes.

So added mediastreamer2 without any patches to guix master as

commit b6b3a6a6c4f00815da576912907c7bdbf02ca11a.

If the patches are required after all (for example the srtp2 one--is it

correct that this one updates the dependency from srtp1 to srtp2?), we

can still add them.

I've also renamed the package to mediastreamer2 like upstream does

(I don't like the name so much--but upstream chose it, so be it).

How come that https://github.com/BelledonneCommunications/liblinphone/releaseshas 4.3.1 and we have 3.12.0 ?

Renamed package to liblinphone and pushed this patch to guix master as

commit 9732910dd336e9bca56991a0630619bc1578d2af.

Pushed to guix master as commit 02b32cdda68495344d8e5521db5aeaf5add03797.

This package is not reproducible. Probably something simple amiss.

diffoscope --exclude-directory-metadata /gnu/store/p9186fvq27dawzmij3q2ibllhdd07c6a-linphoneqt-4.1.1 /gnu/store/p9186fvq27dawzmij3q2ibllhdd07c6a-linphoneqt-4.1.1-check

--- /gnu/store/p9186fvq27dawzmij3q2ibllhdd07c6a-linphoneqt-4.1.1

+++ /gnu/store/p9186fvq27dawzmij3q2ibllhdd07c6a-linphoneqt-4.1.1-check

??? bin

? ??? linphone

? ? ??? /gnu/store/mx2bgrpxkbdjsmhlxp9a30hbzcilk4cn-binutils-2.32/bin/readelf --wide --decompress --hex-dump=.rodata {}

? ? ? @@ -190213,17 +190213,17 @@

? ? ? 0x00801020 02000a00 12000309 3b093c09 3d000109 ........;.<.=...

? ? ? 0x00801030 3e000100 02010c05 56000200 0c000300 >.......V.......

? ? ? 0x00801040 f3023706 03000100 03004c00 4d045700 ..7.......L.M.W.

? ? ? 0x00801050 00000000 00000000 00000000 00000000 ................

? ? ? 0x00801060 00000000 00020000 00010000 00010000 ................

? ? ? 0x00801070 00000000 00000000 00000002 00000002 ................

? ? ? 0x00801080 00000002 00000000 00000000 00000018 ................

? ? ? - 0x00801090 00010000 00010000 00000000 01714bf9 .............qK.

? ? ? - 0x008010a0 65980000 00220001 00000001 0000280f e...."........(.

? ? ? - 0x008010b0 00000171 4bf96598 00000000 00000000 ...qK.e.........

? ? ? + 0x00801090 00010000 00010000 00000000 01714bf7 .............qK.

? ? ? + 0x008010a0 f3330000 00220001 00000001 0000280f .3..."........(.

? ? ? + 0x008010b0 00000171 4bf7f333 00000000 00000000 ...qK..3........

? ? ? 0x008010c0 000904eb 98c3006c 0061006e 00670075 .......l.a.n.g.u

? ? ? 0x008010d0 00610067 00650073 00020000 06be0065 .a.g.e.s.......e

? ? ? 0x008010e0 006e0002 000006d2 00660072 00000000 .n.......f.r....

? ? ? 0x008010f0 00000000 00000000 00000000 00000000 ................

? ? ? 0x00801100 0000280b 00008688 789ce57d 0b741cc5 ..(.....x..}.t..

? ? ? 0x00801110 9568c992 a5d16834 92654996 8531c236 .h....h4.eI..1.6

? ? ? 0x00801120 c626c618 639b9f31 8cf58905 922ceb63 .&..c..1.....,.c

? ??? linphone-tester

? ? ??? /gnu/store/mx2bgrpxkbdjsmhlxp9a30hbzcilk4cn-binutils-2.32/bin/readelf --wide --decompress --hex-dump=.rodata {}

? ? ? @@ -190443,17 +190443,17 @@

? ? ? 0x0080be80 02000a00 12000309 3b093c09 3d000109 ........;.<.=...

? ? ? 0x0080be90 3e000100 02010c05 56000200 0c000300 >.......V.......

? ? ? 0x0080bea0 f3023706 03000100 03004c00 4d045700 ..7.......L.M.W.

? ? ? 0x0080beb0 00000000 00000000 00000000 00000000 ................

? ? ? 0x0080bec0 00000000 00020000 00010000 00010000 ................

? ? ? 0x0080bed0 00000000 00000000 00000002 00000002 ................

? ? ? 0x0080bee0 00000002 00000000 00000000 00000018 ................

? ? ? - 0x0080bef0 00010000 00010000 00000000 01714bf9 .............qK.

? ? ? - 0x0080bf00 65980000 00220001 00000001 0000280f e...."........(.

? ? ? - 0x0080bf10 00000171 4bf96598 00000000 00000000 ...qK.e.........

? ? ? + 0x0080bef0 00010000 00010000 00000000 01714bf7 .............qK.

? ? ? + 0x0080bf00 f3330000 00220001 00000001 0000280f .3..."........(.

? ? ? + 0x0080bf10 00000171 4bf7f333 00000000 00000000 ...qK..3........

? ? ? 0x0080bf20 000904eb 98c3006c 0061006e 00670075 .......l.a.n.g.u

? ? ? 0x0080bf30 00610067 00650073 00020000 06be0065 .a.g.e.s.......e

? ? ? 0x0080bf40 006e0002 000006d2 00660072 00000000 .n.......f.r....

? ? ? 0x0080bf50 00000000 00000000 00000000 00000000 ................

? ? ? 0x0080bf60 0000280b 00008688 789ce57d 0b741cc5 ..(.....x..}.t..

? ? ? 0x0080bf70 9568c992 a5d16834 92654996 8531c236 .h....h4.eI..1.6

? ? ? 0x0080bf80 c626c618 639b9f31 8cf58905 922ceb63 .&..c..1.....,.c

Hi Danny!

Thanks for the push.

Oh no, that patch is required. MS won't build with srtp enabled, without it. You can see "MediaStreamer will be built without SRTP support". That patch fixes search and locate of srtp2 version of include and lib files.

Shall we please keep the name "mediastreamer" for package and "mediastreamer2" for variable? The release tarballs have not been renamed. We'll change the package name when it is changed in releases by upstream. This will consistent with other packages in linphone.scm.

Regards,

RG.

Hi Danny!

Releases tag in gitlab.linphone.org are for internal purposes. Official releases are done at linphone.org/releases/sources.

Hi Danny!

Thanks for the push :-)

Shall we please keep the name "linphone" for package and "liblinphone" for variable? The release tarballs have not been renamed (linphone.org/releases/sources). We'll change the package name when it is changed in releases by upstream. This will consistent with other packages in linphone.scm.

Regards,

RG.

Thanks for the patch!

Pushed to guix master as commit 376fc95e29b24609f3cb7fca024e93a37c0adbfe.

Thanks for the patch!

Pushed to guix master as commit f689d3d7d9ba3fa25b4839bdd1e4b98905f4869c.

Thanks for the patch!

Pushed to guix master as commit fe8f1c0b194de2fa7aa5265a4a0424e82bd11af1.

Hi Raghav,

thanks for the patch!

Pushed to guix master as commit b89201695582351fa5fa183c7a1aa26162ca271e.

Hi Raghav,

On Sun, 05 Apr 2020 22:52:25 +0000

"Raghav Gururajan" <raghavgururajan@disroot.org> wrote:

Okay--added in commit 2c56aeaea94d3914acf79f9bbabeef1b3ec8aa07.

I'll see what the others say--but I'm not opposed to use the other name.

Hi Danny!

Thanks so much :-)

Cool!

Regards,

RG.

Hello,

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

mediastreamer2 is better. Otherwise we get a confused mix of

mediastreamer-2.16.1 / mediastreamer2-2.16.1 under share/, for example.

The name really is mediastreamer2 in the code/doc, everywhere except in

for the tarball file name.

I'll push some refinements I made today in a moment, mostly adding the

tester binary and data to a 'tester' output (it comes with about 20 MiB

of data files).

Great work!

Maxim

Hi Danny!

Nice catch! Let me try to see what causes it.

Btw, is there everything else good with the package?

Regards,

RG.

Hi Danny!

Hmm, I think it is something to with computing QML files.

Hi Raghav,

if we add gcc 5 here, do we need to do something similar to

https://lists.gnu.org/archive/html/guix-commits/2020-04/msg00774.htmltoo or

not?

Why is only gcc 5 working and not gcc 7?

Hi Raghav,

pushed to guix core-updates (with changes below) because, there, it works

without weird special gcc versions.

Commit is c0cce2055757c18619ef1826c54ed3d5b9ec7eb0.

Moved all propagated-inputs to regular inputs--except qttools which I moved to

native-inputs.

Added qtgraphicaleffects to inputs.

Switched to qt build system so the programs actually start up successfully.

There is still the same reproducibility problems in bin/linphone and

bin/linphone-tester:

--- /gnu/store/7g8kdcwzvybx7a8w87pnbypy7x0rf9wg-linphoneqt-4.1.1

+++ /gnu/store/7g8kdcwzvybx7a8w87pnbypy7x0rf9wg-linphoneqt-4.1.1-check

??? bin

? ??? linphone

? ? ??? /gnu/store/mx2bgrpxkbdjsmhlxp9a30hbzcilk4cn-binutils-2.32/bin/readelf --wide --decompress --hex-dump=.rodata {}

? ? ? @@ -190223,17 +190223,17 @@

? ? ? 0x008080c0 02000a00 12000309 3b093c09 3d000109 ........;.<.=...

? ? ? 0x008080d0 3e000100 02010c05 56000200 0c000300 >.......V.......

? ? ? 0x008080e0 f3023706 03000100 03004c00 4d045700 ..7.......L.M.W.

? ? ? 0x008080f0 00000000 00000000 00000000 00000000 ................

? ? ? 0x00808100 00000000 00020000 00010000 00010000 ................

? ? ? 0x00808110 00000000 00000000 00000002 00000002 ................

? ? ? 0x00808120 00000002 00000000 00000000 00000018 ................

? ? ? - 0x00808130 00010000 00010000 00000000 0171666d .............qfm

? ? ? - 0x00808140 dfb80000 00220001 00000001 0000280f ....."........(.

? ? ? - 0x00808150 00000171 666ddfb8 00000000 00000000 ...qfm..........

? ? ? + 0x00808130 00010000 00010000 00000000 0171666c .............qfl

? ? ? + 0x00808140 71bb0000 00220001 00000001 0000280f q...."........(.

? ? ? + 0x00808150 00000171 666c71bb 00000000 00000000 ...qflq.........

? ? ? 0x00808160 000904eb 98c3006c 0061006e 00670075 .......l.a.n.g.u

? ? ? 0x00808170 00610067 00650073 00020000 06be0065 .a.g.e.s.......e

? ? ? 0x00808180 006e0002 000006d2 00660072 00000000 .n.......f.r....

? ? ? 0x00808190 00000000 00000000 00000000 00000000 ................

? ? ? 0x008081a0 0000280b 00008688 789ce57d 0b741cc5 ..(.....x..}.t..

? ? ? 0x008081b0 9568c992 a5d16834 92654996 8531c236 .h....h4.eI..1.6

? ? ? 0x008081c0 c626c618 639b9f31 8cf58905 922ceb63 .&..c..1.....,.c

? ??? linphone-tester

? ? ??? /gnu/store/mx2bgrpxkbdjsmhlxp9a30hbzcilk4cn-binutils-2.32/bin/readelf --wide --decompress --hex-dump=.rodata {}

? ? ? @@ -190455,17 +190455,17 @@

? ? ? 0x00811f40 02000a00 12000309 3b093c09 3d000109 ........;.<.=...

? ? ? 0x00811f50 3e000100 02010c05 56000200 0c000300 >.......V.......

? ? ? 0x00811f60 f3023706 03000100 03004c00 4d045700 ..7.......L.M.W.

? ? ? 0x00811f70 00000000 00000000 00000000 00000000 ................

? ? ? 0x00811f80 00000000 00020000 00010000 00010000 ................

? ? ? 0x00811f90 00000000 00000000 00000002 00000002 ................

? ? ? 0x00811fa0 00000002 00000000 00000000 00000018 ................

? ? ? - 0x00811fb0 00010000 00010000 00000000 0171666d .............qfm

? ? ? - 0x00811fc0 dfb80000 00220001 00000001 0000280f ....."........(.

? ? ? - 0x00811fd0 00000171 666ddfb8 00000000 00000000 ...qfm..........

? ? ? + 0x00811fb0 00010000 00010000 00000000 0171666c .............qfl

? ? ? + 0x00811fc0 71bb0000 00220001 00000001 0000280f q...."........(.

? ? ? + 0x00811fd0 00000171 666c71bb 00000000 00000000 ...qflq.........

? ? ? + 0x00811fb0 00010000 00010000 00000000 0171666c .............qfl

? ? ? + 0x00811fc0 71bb0000 00220001 00000001 0000280f q...."........(.

? ? ? + 0x00811fd0 00000171 666c71bb 00000000 00000000 ...qflq.........

? ? ? 0x00811fe0 000904eb 98c3006c 0061006e 00670075 .......l.a.n.g.u

? ? ? 0x00811ff0 00610067 00650073 00020000 06be0065 .a.g.e.s.......e

? ? ? 0x00812000 006e0002 000006d2 00660072 00000000 .n.......f.r....

? ? ? 0x00812010 00000000 00000000 00000000 00000000 ................

? ? ? 0x00812020 0000280b 00008688 789ce57d 0b741cc5 ..(.....x..}.t..

? ? ? 0x00812030 9568c992 a5d16834 92654996 8531c236 .h....h4.eI..1.6

? ? ? 0x00812040 c626c618 639b9f31 8cf58905 922ceb63 .&..c..1.....,.c

Hi Danny!

The requirement for gcc 5 is related to Qt. It has something to do with C++ standard. The Qt package in qt.scm uses gcc 5 instead of gcc 7.

The reason for gcc 7 working for linphoneqt on core-updates, could also be due to updates in qt.

Regards,

RG.

Hello Raghav!

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

To have a mirror makes sense when many locations provide the same

content. Here there's only https://www.linphone.org/so I don't think

adding a mirror for it is necessary.

Thank you,

Maxim

Hello Danny,

Danny Milosavljevic <dannym@scratchpost.org> writes:

In my case I'm seeing slightly different outputs:

diffoscope --exclude-command stat /gnu/store/i3pqi987rdm3x023ya4zd2k3jxhaiqqy-linphoneqt-4.1.1{,-check}

--- /gnu/store/i3pqi987rdm3x023ya4zd2k3jxhaiqqy-linphoneqt-4.1.1

+++ /gnu/store/i3pqi987rdm3x023ya4zd2k3jxhaiqqy-linphoneqt-4.1.1-check

??? bin

? ??? .linphone-real

? ? ??? /gnu/store/mx2bgrpxkbdjsmhlxp9a30hbzcilk4cn-binutils-2.32/bin/readelf --wide --decompress --hex-dump=.rodata {}

? ? ? @@ -190223,17 +190223,17 @@

? ? ? 0x008080c0 02000a00 12000309 3b093c09 3d000109 ........;.<.=...

? ? ? 0x008080d0 3e000100 02010c05 56000200 0c000300 >.......V.......

? ? ? 0x008080e0 f3023706 03000100 03004c00 4d045700 ..7.......L.M.W.

? ? ? 0x008080f0 00000000 00000000 00000000 00000000 ................

? ? ? 0x00808100 00000000 00020000 00010000 00010000 ................

? ? ? 0x00808110 00000000 00000000 00000002 00000002 ................

? ? ? 0x00808120 00000002 00000000 00000000 00000018 ................

? ? ? - 0x00808130 00010000 00010000 00000000 01717ba8 .............q{.

? ? ? - 0x00808140 1fbe0000 00220001 00000001 0000280f ....."........(.

? ? ? - 0x00808150 00000171 7ba81fba 00000000 00000000 ...q{...........

? ? ? + 0x00808130 00010000 00010000 00000000 01717ba6 .............q{.

? ? ? + 0x00808140 59760000 00220001 00000001 0000280f Yv..."........(.

? ? ? + 0x00808150 00000171 7ba65972 00000000 00000000 ...q{.Yr........

? ? ? 0x00808160 000904eb 98c3006c 0061006e 00670075 .......l.a.n.g.u

? ? ? 0x00808170 00610067 00650073 00020000 06be0065 .a.g.e.s.......e

? ? ? 0x00808180 006e0002 000006d2 00660072 00000000 .n.......f.r....

? ? ? 0x00808190 00000000 00000000 00000000 00000000 ................

? ? ? 0x008081a0 0000280b 00008688 789ce57d 0b741cc5 ..(.....x..}.t..

? ? ? 0x008081b0 9568c992 a5d16834 92654996 8531c236 .h....h4.eI..1.6

? ? ? 0x008081c0 c626c618 639b9f31 8cf58905 922ceb63 .&..c..1.....,.c

? ??? .linphone-tester-real

? ? ??? /gnu/store/mx2bgrpxkbdjsmhlxp9a30hbzcilk4cn-binutils-2.32/bin/readelf --wide --decompress --hex-dump=.rodata {}

? ? ? @@ -190455,17 +190455,17 @@

? ? ? 0x00811f40 02000a00 12000309 3b093c09 3d000109 ........;.<.=...

? ? ? 0x00811f50 3e000100 02010c05 56000200 0c000300 >.......V.......

? ? ? 0x00811f60 f3023706 03000100 03004c00 4d045700 ..7.......L.M.W.

? ? ? 0x00811f70 00000000 00000000 00000000 00000000 ................

? ? ? 0x00811f80 00000000 00020000 00010000 00010000 ................

? ? ? 0x00811f90 00000000 00000000 00000002 00000002 ................

? ? ? 0x00811fa0 00000002 00000000 00000000 00000018 ................

? ? ? - 0x00811fb0 00010000 00010000 00000000 01717ba8 .............q{.

? ? ? - 0x00811fc0 1fbe0000 00220001 00000001 0000280f ....."........(.

? ? ? - 0x00811fd0 00000171 7ba81fba 00000000 00000000 ...q{...........

? ? ? + 0x00811fb0 00010000 00010000 00000000 01717ba6 .............q{.

? ? ? + 0x00811fc0 59760000 00220001 00000001 0000280f Yv..."........(.

? ? ? + 0x00811fd0 00000171 7ba65972 00000000 00000000 ...q{.Yr........

? ? ? 0x00811fe0 000904eb 98c3006c 0061006e 00670075 .......l.a.n.g.u

? ? ? 0x00811ff0 00610067 00650073 00020000 06be0065 .a.g.e.s.......e

? ? ? 0x00812000 006e0002 000006d2 00660072 00000000 .n.......f.r....

? ? ? 0x00812010 00000000 00000000 00000000 00000000 ................

? ? ? 0x00812020 0000280b 00008688 789ce57d 0b741cc5 ..(.....x..}.t..

? ? ? 0x00812030 9568c992 a5d16834 92654996 8531c236 .h....h4.eI..1.6

? ? ? 0x00812040 c626c618 639b9f31 8cf58905 922ceb63 .&..c..1.....,.c

Adding more data points so that hopefully we can find out the cause of

the non-reproducibility issue.

Maxim

Hello again,

I've isolated the culprit: it's the translation auto-generated files

qrc_i18n.cpp. I've run a couple builds locally, then diffed the build

files, and found this suspicious:

diff -I build2 -I build3 -r build2/qrc_i18n.cpp build3/qrc_i18n.cpp

1352c1352

< 0x0,0x0,0x1,0x71,0x7b,0xea,0x68,0x8a,

---

1355c1355

< 0x0,0x0,0x1,0x71,0x7b,0xea,0x68,0x86,

---

I then compared the object files for this source, e.g.: diffoscope

--exclude-command stat

build{2,3}/CMakeFiles/app-library.dir/qrc_i18n.cpp.o

Hex dump of section '.rodata':

? 0x00000000 00000000 00020000 00010000 00010000 ................

? 0x00000010 00000000 00000000 00000002 00000002 ................

? 0x00000020 00000002 00000000 00000000 00000018 ................

? - 0x00000030 00010000 00010000 00000000 01717bea .............q{.

? - 0x00000040 688a0000 00220001 00000001 0000280f h...."........(.

? - 0x00000050 00000171 7bea6886 00000000 00000000 ...q{.h.........

? + 0x00000030 00010000 00010000 00000000 01717bec .............q{.

? + 0x00000040 98610000 00220001 00000001 0000280f .a..."........(.

? + 0x00000050 00000171 7bec9861 00000000 00000000 ...q{..a........

? 0x00000060 000904eb 98c3006c 0061006e 00670075 .......l.a.n.g.u

? 0x00000070 00610067 00650073 00020000 06be0065 .a.g.e.s.......e

? 0x00000080 006e0002 000006d2 00660072 00000000 .n.......f.r....

? 0x00000090 00000000 00000000 00000000 00000000 ................

? 0x000000a0 0000280b 00008688 789ce57d 0b741cc5 ..(.....x..}.t..

? 0x000000b0 9568c992 a5d16834 92654996 8531c236 .h....h4.eI..1.6

? 0x000000c0 c626c618 639b9f31 8cf58905 922ceb63 .&..c..1.....,.c

This matches the .rodata diff found in the main binary :-).

Maxim

Hi!

Maxim Cournoyer <maxim.cournoyer@gmail.com> writes:

Fixed in commit a4f1a7e2db, on core-updates :-)

Maxim

Hello Raghav,

"Raghav Gururajan" <raghavgururajan@disroot.org> writes:

Please not that we usually submit changes *per-package*, that is, one

commit per package. It's a convention we use. I'm making an exception

here because the changes are cosmetic only, but keep this in mind for

your next changes.

I've pushed a slightly modified version of the above changes in commit

1988ef21f82430d4b380dd4887160a902d8796fd (on the core-updates branch).

Thank you for your contributions and your patience in the lengthy review

process.

Maxim