Your message dated Sun, 21 Jul 2024 13:34:44 +0200
with message-id <zpzyvi1vvvt0f...@aurel32.net>
and subject line Re: Bug#1073859: gopacket: FTBFS with glibc 2.39 due to
internal type __time64_t
has caused the Debian Bug report #1073859,
regarding gopacket: FTBFS with glibc 2.39 due to internal type __time64_t
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact ow...@bugs.debian.org
immediately.)
--
1073859: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1073859
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
--- Begin Message ---
Source: gopacket
Version: 1.1.19-6
Severity: important
Tags: patch ftbfs
User: debian-gl...@lists.debian.org
Usertags: glibc2.39
Control: affects -1 bettercap
Dear maintainer,
gopacket fail to build from source when built against glibc 2.39:
| src/github.com/google/gopacket/pcap/pcap_unix.go:349:18: could not determine
kind of name for C.gopacket_time_secs_t
Packages depending on golang-github-google-gopacket-dev are also
affected.
It appears that the issue has been introduced in version 1.1.19-5 by
0002-fix-time-type-with-__USE_TIME_BITS64.patch. It uses __time64_t when
__USE_TIME_BITS64 is defined, but __time64_t is an internal type which
must never used by user code. Among other things it is not defined in
when it equals __time_t. glibc 2.39 changed the cases where
__USE_TIME_BITS64 is defined.
The easiest way to fix that is to just drop the glibc specific part and
use the generic definition, which relies on correct definition of time_t
and suseconds_t, and avoid logic duplication between gopacket and glibc.
The following patch does that and could be used as a replacement for
0002-fix-time-type-with-__USE_TIME_BITS64.patch:
--- a/pcap/pcap_unix.go
+++ b/pcap/pcap_unix.go
@@ -111,9 +111,6 @@
#elif __ANDROID__
#define gopacket_time_secs_t __kernel_time_t
#define gopacket_time_usecs_t __kernel_suseconds_t
-#elif __GLIBC__
-#define gopacket_time_secs_t __time_t
-#define gopacket_time_usecs_t __suseconds_t
#else // Some form of linux/bsd/etc...
#include <sys/param.h>
#ifdef __OpenBSD__
Regards
Aurelien
--- End Message ---
--- Begin Message ---
Version: 1.1.19-6.2
On 2024-07-19 23:37, Aurelien Jarno wrote:
> Hi,
>
> On 2024-07-18 01:15, Reinhard Tartler wrote:
> > Aurelien Jarno <aure...@debian.org> writes:
> >
> > > I have just uploaded a NMU to delayed/2 using the above strategy. Please
> > > feel free to ask me to delay or cancel it. You will find the
> > > corresponding debdiff attached.
> >
> > I've looked over the patch, it LGTM. Thank you for providing the patch!
> >
> > I understand it is going to land later today, that's awesome! If for
> > some reason it should fall throught he cracks, please upload it directly
> > to sid.
>
> It appears that the fix was not fully correct and trigger an autopkgtest
> regression on armhf and armel. I took the freedom to upload a fix,
> please find the debdiff attached.
>
This timed is worked, with no autopkgtest regression. Closing the bug
accordingly.
Regards
Aurelien
--
Aurelien Jarno GPG: 4096R/1DDD8C9B
aurel...@aurel32.net http://aurel32.net
--- End Message ---
