Builder libpcap-solaris10-amd64 Build #3236

Results:

Build successful

SourceStamp:

BuildSlave:

Reason:

The SingleBranchScheduler scheduler named 'schedule-libpcap-solaris10-amd64' triggered this build

Steps and Logfiles:

1. git update ( 7 secs )
   1. stdio
2. shell Set developer mode ( 0 secs )
   1. stdio
3. shell_1 './autogen.sh' ( 6 secs )
   1. stdio
4. shell_2 './configure' ( 53 secs )
   1. stdio
   2. config.log
5. shell_3 'gmake' ( 27 secs )
   1. stdio
6. shell_4 'gmake check' ( 15 mins, 16 secs )
   1. stdio
7. shell_5 'rm -rf ...' ( 0 secs )
   1. stdio
8. shell_6 'gmake install ...' ( 3 secs )
   1. stdio
9. trigger triggered trigger-tcpdump-solaris10-amd64 ( 6 mins, 59 secs )
   1. - no logs -
   2. tcpdump-solaris10-amd64 #6060

Build Properties:

Forced Build Properties:

Responsible Users:

1. Denis Ovsienko

   denisohnoyoudont@ovsienko.info

Timing:

All Changes:

:

1. Change #256436

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:19:48
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	23139e3f6badd02e9469cefc325cbfa0e9d7fcf8

   Comments

   Put a few comments right.
   In gen_linktype() remove a note that went stale after commit a8f23f7.

   Changed files

   • gencode.c
   • testprogs/capturetest.c

2. Change #256437

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:19:48
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	76b3026726ea670f9952ab88b15b05cbf9d46746

   Comments

   Factor gen_ip_version() out.

   Changed files

   • gencode.c

3. Change #256438

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:19:48
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	2de388f4b76c14fcae5adf75555df552302ec61c

   Comments

   Introduce named constants for MPLS label stack.

   Changed files

   • gencode.c

4. Change #256439

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:19:48
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	287874e74dc48d4937c1f9460a38be05490bcdf2

   Comments

   Refine MPLS Bottom of Stack bit matching.
   As can be seen in gen_load_a(), for OR_LINKPL offset -2 is the same as
   offset 2 for OR_PREVMPLSHDR, so both gen_mpls_linktype() and
   gen_mpls_internal() test the same single bit, except for different
   values.
   
   Make the bit set test a new function and there produce "jset #0x1"
   instead of "and #0x1; jeq #0x1": this branches the same way, but uses
   one fewer instruction and preserves the accumulator value after the
   load.  For the bit clear test use the same with gen_not() around it.
   Update the tests.

   Changed files

   • gencode.c
   • testprogs/TESTrun

5. Change #256440

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:19:48
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	e43e2ede39b37d3887ccbbc98fb715cbffa9994c

   Comments

   Check for <machine/atomic.h> after all module auto-detection.
   In commit fc3781e I added the header check into the BPF auto-detection
   step, which skips the check when the user specifies the module
   explicitly, which disables zero-copy when BPF is the module:
   
   $ uname -v
   FreeBSD 14.3-RELEASE-p7 GENERIC
   
   $ ./configure -q
   $ grep MACHINE_ATOMIC_H config.h
   #define HAVE_MACHINE_ATOMIC_H 1
   $ make -s testprogs && ./testprogs/versiontest
   libpcap version 1.11.0-PRE-GIT (64-bit time_t, with zerocopy and netmap
     support)
   
   $ ./configure -q --with-pcap=bpf
   $ grep MACHINE_ATOMIC_H config.h
   /* #undef HAVE_MACHINE_ATOMIC_H */
   $ make -s testprogs && ./testprogs/versiontest
   libpcap version 1.11.0-PRE-GIT (64-bit time_t, with netmap support)
   
   Move the check to a module-specific step later on to fix this:
   
   $ ./configure -q --with-pcap=bpf
   $ grep MACHINE_ATOMIC_H config.h
   #define HAVE_MACHINE_ATOMIC_H 1
   $ make -s testprogs && ./testprogs/versiontest
   libpcap version 1.11.0-PRE-GIT (64-bit time_t, with zerocopy and netmap
     support)

   Changed files

   • CMakeLists.txt
   • configure.ac

6. Change #256441

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:19:48
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	ece234eef1303fd06b03cdc969cbf2601af14066

   Comments

   Check for the __atomic builtins on Linux only.
   Only pcap-linux.c needs to know whether the builtins are available, so
   do the checks only when the file is going to be compiled.

   Changed files

   • CMakeLists.txt
   • configure.ac

7. Change #256442

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:19:48
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	7dc97ac2a3fd6ed9e3c6df1efb8031a3e0b41231

   Comments

   man: Format two em dashes properly.
   roff provides a number of different ways to spell an em dash.  The
   oldest and the most universally supported syntax seems to be "\(em",
   which is what tcpdump(1) has been using since at least tcpdump 2.0.

   Changed files

   • pcap-filter.manmisc.in
   • pcap_findalldevs.3pcap

8. Change #256443

   Category	libpcap
   Changed by	Denis Ovsienko <denisohnoyoudont@ovsienko.info>
   Changed at	Mon 26 Jan 2026 19:20:08
   Repository	https://git.tcpdump.org/libpcap
   Project	libpcap
   Branch	master
   Revision	92bf5de12e5e6b6509b3158000a252dfae8c9970

   Comments

   pcap-filter(7): Define MAC-48 formats precisely. [skip ci]

   Changed files

   • pcap-filter.manmisc.in