Builder libpcap-solaris10-sparcv9 Build #3111
Results:
Build successful
SourceStamp:
Project | libpcap |
Repository | https://git.tcpdump.org/libpcap |
Branch | master |
Revision | ca63110cee408812bec55d322ed500cdb3f4e58b |
Got Revision | ca63110cee408812bec55d322ed500cdb3f4e58b |
Changes | 3 changes |
BuildSlave:
unstable10sReason:
The SingleBranchScheduler scheduler named 'schedule-libpcap-solaris10-sparcv9' triggered this build
Steps and Logfiles:
-
git update ( 8 secs )
-
shell Set developer mode ( 0 secs )
-
shell_1 './autogen.sh' ( 20 secs )
-
shell_2 './configure' ( 1 mins, 2 secs )
-
shell_3 'gmake' ( 48 secs )
-
shell_4 'rm -rf ...' ( 0 secs )
-
shell_5 'gmake install ...' ( 4 secs )
-
trigger triggered trigger-tcpdump-solaris10-sparcv9 ( 11 mins, 5 secs )
- - no logs -
- tcpdump-solaris10-sparcv9 #5865
Build Properties:
Name | Value | Source |
---|---|---|
branch | master | Build |
builddir | /export/home/buildbot-unstable10s/slave/libpcap-solaris10-sparcv9 | slave |
buildername | libpcap-solaris10-sparcv9 | Builder |
buildnumber | 3111 | Build |
codebase | Build | |
got_revision | ca63110cee408812bec55d322ed500cdb3f4e58b | Git |
project | libpcap | Build |
repository | https://git.tcpdump.org/libpcap | Build |
revision | ca63110cee408812bec55d322ed500cdb3f4e58b | Build |
scheduler | schedule-libpcap-solaris10-sparcv9 | Scheduler |
slavename | unstable10s | BuildSlave |
workdir | /export/home/buildbot-unstable10s/slave/libpcap-solaris10-sparcv9 | slave (deprecated) |
Forced Build Properties:
Name | Label | Value |
---|
Responsible Users:
- Denis Ovsienkodenis@ovsienko.info
Timing:
Start | Thu Feb 27 21:50:26 2025 |
End | Thu Feb 27 22:03:56 2025 |
Elapsed | 13 mins, 30 secs |
All Changes:
:
Change #223324
Category libpcap Changed by Denis Ovsienko <denis @ovsienko.info>Changed at Wed 26 Feb 2025 15:16:45 Repository https://git.tcpdump.org/libpcap Project libpcap Branch master Revision 66941d599aa6e93754f10dd9404715aaa345ac5b Comments
TESTrun: Add apply filter tests for DLT_SLIP. Capture the current behaviour of "inbound" and "outbound" applied to Rx, Tx and invalid packet direction. Note how "outbound" matches a packet with an invalid direction value (0xE7). The .pcap files come from tcpdump tests.
Changed files
- testprogs/TESTrun
- tests/filter/slip-bad-direction.pcap
- tests/filter/slip-compressed_sl_print-oobr.pcap
- tests/filter/slip-sliplink_print-oobr.pcap
Change #223325
Category libpcap Changed by Denis Ovsienko <denis @ovsienko.info>Changed at Thu 27 Feb 2025 21:34:08 Repository https://git.tcpdump.org/libpcap Project libpcap Branch master Revision 127435bf8c1e3eb08166a6004b74d638e5afd4fb Comments
Have "outbound" mean Tx only for DLT_SLIP. In DLT_SLIP packet direction is the first and the only byte of the link-layer header, so it ought to be trivial to test, which the optimized filter program seems to implement: $ ./testprogs/filtertest SLIP outbound (000) ldb [0] (001) jeq #0x0 jt 2 jf 3 (002) ret #0 (003) ret #262144 However, the C code in gen_inbound_outbound() looks more complicated than the above requires: since libpcap 0.0 (June 1994) it has been a combination of four functions, and the unoptimized filter program it produces looks more complicated than it needs to be: $ ./testprogs/filtertest -O SLIP outbound (000) ld #0x0 (001) st M[1] (002) ldx M[1] (003) ldb [x + 0] (004) st M[2] (005) ld #0x0 (006) st M[0] (007) ldx M[0] (008) ld M[2] (009) sub x (010) jeq #0x0 jt 11 jf 12 (011) ret #0 (012) ret #262144 The C code means loading 1 byte from offset 0 of the link-layer header, testing it to be equal to 0 and reversing the result if testing for "outbound". The test for "inbound" correctly matches direction value 0 (received packets). The test for "outbound" matches direction value 1 (transmitted packets), as well as any other non-zero direction value, which is invalid in DLT_SLIP. Replace the custom code with gen_cmp() consistently with the other cases in the same switch block. This makes the unoptimized and the optimized filter programs identical, also fixes "outbound" to match Tx packets only. Update the affected filter tests to reflect these developments.
Changed files
- CHANGES
- gencode.c
- testprogs/TESTrun
Change #223326
Category libpcap Changed by Denis Ovsienko <denis @ovsienko.info>Changed at Thu 27 Feb 2025 21:34:37 Repository https://git.tcpdump.org/libpcap Project libpcap Branch master Revision ca63110cee408812bec55d322ed500cdb3f4e58b Comments
pcap-filter(7): Refine qualifiers and primitives. In the introduction mention that an ID is not limited to a name or a number, add "proto" to the list of type qualifiers and give an example, mention that a proto qualifier and the "proto" type qualifier are not the same, add more entries to the list of valid protocols, give more examples, say that "proto" implies "ip or ip6" and "net" also implies "ip6". Ibid., add a paragraph to explain how the qualifiers combine (or not) in the regular primitive syntax; in IEEE 802.11 header particulars remove the comment about RA and TA because the current implementation does in fact support these, remove the comment about BSSID because the list of "wlan" primitives represents what is supported, everything else is not. Ibid., add ARP/RARP header particulars, rearrange a couple paragraphs to discuss one thing at a time and provide a few more examples of the irregular primitive syntax. For each of the primitives below merge "dst", "src" and "src or dst" forms into one section using a uniform style and in each section give a complete list of valid qualifiers and related details: * "host HOSTNAMEADDR" * "ether host ETHERNAMEADDR" * "net NETNAMEADDR" * "net NETADDR mask NETMASK" * "net NETADDR/LEN" * "port PORTNAMENUM" * "portrange PORTNAMENUM1-PORTNAMENUM2" * "decnet HOST" * "link HOST" with regard to ARCnet Update "gateway HOST" along the same lines to make it clear it matches ARP and RARP too by default. For "ether host ETHERNAMEADDR", "ether broadcast" and "ether multicast" mention that Ethernet is not the only supported network type. For "net NETNAMEADDR" spell the CIDR notation of every netmask. Ibid., add a reference to "net NETADDR/LEN". Idem for "net NETADDR mask NETMASK", also discuss the argument syntax and clarify that the netmask can be any 32-bit value. For "portrange" make it clear that the two values can be in any order and may be equal. Add a note about ARP and RARP to the "bugs" section.
Changed files
- CHANGES
- pcap-filter.manmisc.in