Bug 247914 - net-p2p/gtk-gnutella - Update to 1.2.0
Summary: net-p2p/gtk-gnutella - Update to 1.2.0
Status: New
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: Any Any
: --- Affects Some People
Assignee: Kurt Jaeger
URL:
Keywords: buildisok
Depends on:
Blocks:
 
Reported: 2020-07-11 18:51 UTC by Jamie Landeg-Jones
Modified: 2020-09-26 07:19 UTC (History)
3 users (show)

See Also:
bugzilla: maintainer-feedback? (woodsb02)


Attachments
cd /usr/ports/net-p2p/gtk-gnutella && patch < <filename> (1.17 KB, patch)
2020-07-11 18:51 UTC, Jamie Landeg-Jones
no flags Details | Diff
New patch. Obsoletes previous one. patch -p0 < .... (6.11 KB, patch)
2020-09-26 07:19 UTC, Jamie Landeg-Jones
no flags Details | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Jamie Landeg-Jones 2020-07-11 18:51:59 UTC
Created attachment 216384 [details]
cd /usr/ports/net-p2p/gtk-gnutella && patch < <filename>

Patch for update to latest version.

Note, new build requires "gmake", other than that, a pretty bog standard update from a ports point of view

Cheers, Jamie
Comment 1 Automation User 2020-07-26 00:39:51 UTC
Build info is available at https://gitlab.com/swills/freebsd-ports/pipelines/170786666
Comment 2 Jamie Landeg-Jones 2020-08-31 20:36:45 UTC
I'll take ownership if this is abandoned...
Comment 3 Kurt Jaeger freebsd_committer 2020-09-01 09:06:34 UTC
testbuilds@work
Comment 4 Kurt Jaeger freebsd_committer 2020-09-01 12:00:31 UTC
It builds on amd64, but fails on current-i386, see

https://people.freebsd.org/~pi/logs/net-p2p__gtk-gnutella-curi-1598951181.txt

Do you want to investigate or is i386 no longer supported ?
Comment 5 Jamie Landeg-Jones 2020-09-02 22:25:19 UTC
Sorry about that. I'll investigate, and speak to the author.

Cheers, Jamie
Comment 6 Ben Woods freebsd_committer 2020-09-21 23:00:24 UTC
Hi Jamie,

How did you go - get any feedback on why gtk-gnutella has stopped building on i386 with this update?

Are you able to provide a link to an upstream bug report?

Do you think the preference is to update to the latest release and mark i386 as broken, or hold off on the update?

Cheers,
Ben
Comment 7 Jamie Landeg-Jones 2020-09-26 07:18:39 UTC
Hi Ben. I include a new patch file to replace the previous one.

I don't have an upstream bug report - I've been talking to Raphael (the author) directly.

Anyway, I found the problem, and he has confirmed my fix is correct. Basically he'd made a change and missed off i386 related settings for when the system malloc is being used.

The patch will be merged upstream, and is included here.

This wasn't noticed because he develops (and tests extensively on FreeBSD amd64) using his own recommended malloc implementation.

I'm not sure why the default setting for this port was to use system malloc, but if there are no objections, I've changed the port to now use his internal malloc by default, but added an OPTION tag so that it is configurable via the config dialog.

I've also updated and tidied up pkg-help.

Note, I needed to use patch -p0 < ...   to apply this (and then remove the zero length files/patch-src_lib_xmalloc.c.orig manually) - if you know of a better way to create the patch file, I'm all ears! :-)

Let me know if this is ok!

Cheers, Jamie
Comment 8 Jamie Landeg-Jones 2020-09-26 07:19:52 UTC
Created attachment 218312 [details]
New patch. Obsoletes previous one. patch -p0 < ....