On Mon, Oct 15, 2012 at 7:14 PM, <meino.cra...@gmx.de> wrote: > Hi, > > My environment: > Two computers: > a) Desktop PC running Gentoo Linux on a AMD Phenom(tm) II X6 1090T > Processor > b) A beaglebone single board computer running Gentoo Linux on a ARMv7 > Processor rev 2 (v7l) > > a) has the IP-address 192.168.178.21 > b) has the IP-address 192.168.178.25 > > The task: > I want to emerge packages (that is compiling them from source by > using Gentoo package manager) via "pump emerge". > > The problem: > It does not work.... ;/ > > I followed different howtos from different parts of this century ;) > googled from the net how to accomplish this...but no luck. > Often there was no clear distinction between hot and server. > I also reinstalled distcc several times. > > My current setup is as follows: > I build a cross compiler toolchain on a), the server, named > armv7a-softfp-linux-gnueabi-* > > On the beaglebone > > ls -l /usr/lib/distcc/bin > > shows > > lrwxrwxrwx 1 root root 15 2012-10-15 19:21 > armv7a-softfp-linux-gnueabi-c++ -> /usr/bin/distcc > lrwxrwxrwx 1 root root 15 2012-10-15 19:21 > armv7a-softfp-linux-gnueabi-g++ -> /usr/bin/distcc > lrwxrwxrwx 1 root root 15 2012-10-15 19:21 > armv7a-softfp-linux-gnueabi-gcc -> /usr/bin/distcc > -rwxr-xr-x 1 root root 83 2012-10-14 15:11 > armv7a-softfp-linux-gnueabi-wrapper > lrwxrwxrwx 1 root root 35 2012-10-15 19:38 c++ -> > armv7a-softfp-linux-gnueabi-wrapper > lrwxrwxrwx 1 root root 35 2012-10-15 19:39 cc -> > armv7a-softfp-linux-gnueabi-wrapper > lrwxrwxrwx 1 root root 35 2012-10-15 19:39 g++ -> > armv7a-softfp-linux-gnueabi-wrapper > lrwxrwxrwx 1 root root 35 2012-10-15 19:39 gcc -> > armv7a-softfp-linux-gnueabi-wrapper > > and > > armv7a-softfp-linux-gnueabi-wrapper > > has the contents: > > #!/bin/bash > exec /usr/lib/distcc/bin/armv7a-softfp-linux-gnueabi-g${0:$[-2]} "$@" > > > On that machine, the server I started distccd with > > /usr/bin/distccd --daemon --pid-file /var/run/distccd/distccd.pid > --port 3632 --log-file=/tmp/NFS/distcc.solfire.log --log-level critical > --allow 192.168.178.25 -N 15 > > (the logfile is placed in a directory, which is nfs-mounted on machine > "b" so I have access to the log from both nmachines) > > > On the beaglebone, machine "b", the client, the command > > distcc-config -get-host > > will print > > 192.168.178.21,cpp,lzo > > and > > distcc-config --get-log > > will print > > DISTCC_LOG=/tmp/distcc.log > > Then I wrote a little C-program called test.c and did a > > pump make test > > , which lead to the following contents of distcc.log: > > distcc[15967] (dcc_trace_version) distcc 3.1 armv7a-softfp-linux-gnueabi; > built Oct 15 2012 19:19:29 > distcc[15967] (dcc_recursion_safeguard) safeguard level=0 > distcc[15967] (main) compiler name is "distcc" > distcc[15967] (dcc_get_hostlist) not reading /root/.distcc/hosts: No such > file or directory > distcc[15967] (dcc_parse_hosts_file) load hosts from /etc/distcc/hosts > distcc[15967] (dcc_parse_hosts) found tcp token "192.168.178.21,cpp,lzo" > distcc[15967] (dcc_parse_options) got CPP option > distcc[15967] (dcc_parse_options) got LZO option > distcc[15967] (dcc_exit) exit: code 0; self: 0.000000 user 0.010000 sys; > children: 0.000000 user 0.000000 sys >
The lines above indicate that you don't have "localhost" in your ,cpp,zlo" > distcc[15967] (dcc_cleanup_tempfiles_inner) deleted 0 temporary files > distcc[15990] (dcc_trace_version) distcc 3.1 armv7a-softfp-linux-gnueabi; > built Oct 15 2012 19:19:29 > distcc[15990] (dcc_recursion_safeguard) safeguard level=0 > distcc[15990] (main) compiler name is "armv7a-softfp-linux-gnueabi-gcc" > distcc[15990] (dcc_set_path) setting > PATH=/usr/lib/distcc/bin:/usr/local/bin:/bin/:/sbin:/usr/sbin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/opt/bin:/usr/armv7a-softfp-linux-gnueabi/gcc-bin/4.5.4:/root/bin:/root/bin > distcc[15990] (dcc_scan_args) scanning arguments: > armv7a-softfp-linux-gnueabi-gcc -O2 -pipe -march=armv7-a -mfpu=vfpv3-d16 > -mfloat-abi=softfp test.c -o test > distcc[15990] (dcc_scan_args) found input file "test.c" > distcc[15990] (dcc_scan_args) found object/output file "test" > distcc[15990] (dcc_scan_args) compiler apparently called not for compile > That is the critical line which explains why this particular compiler invocation was not distributed to a remote host. Your makefile ran the command "$(CC) $(CFLAGS) test.c -o test", which does two things: compilation *and* linking. Distcc does not attempt to distribute linking, it will only distribute compilation. If you do compilation and linking in a single command, distcc will not attempt to distribute that. Instead of using a single command that combines compilation and linking $(CC) $(CFLAGS) test.c -o test you should use separate commands $(CC) $(CFLAGS) -c test.c $(CC) $(CFLAGS) test.o -o test and then distcc should be able to distribute the compilation command (the first of those two). distcc[15990] (dcc_lock_host) got cpu lock on localhost slot 0 as fd4 > distcc[15990] exec on localhost: armv7a-softfp-linux-gnueabi-gcc -O2 -pipe > -march=armv7-a -mfpu=vfpv3-d16 -mfloat-abi=softfp test.c -o test > distcc[15990] (dcc_note_state) note state 5, file "test.c", host > "localhost" > distcc[15990] (dcc_spawn_child) forking to execute: > armv7a-softfp-linux-gnueabi-gcc -O2 -pipe -march=armv7-a -mfpu=vfpv3-d16 > -mfloat-abi=softfp test.c -o test > distcc[15990] (dcc_spawn_child) child started as pid15991 > distcc[15991] (dcc_increment_safeguard) setting safeguard: > _DISTCC_SAFEGUARD=1 > distcc[15991] (dcc_trace_version) distcc 3.1 armv7a-softfp-linux-gnueabi; > built Oct 15 2012 19:19:29 > distcc[15991] (dcc_recursion_safeguard) safeguard: 1 > distcc[15991] (dcc_recursion_safeguard) safeguard level=1 > distcc[15991] (main) compiler name is "armv7a-softfp-linux-gnueabi-gcc" > distcc[15991] (dcc_set_path) setting > PATH=/usr/local/bin:/bin/:/sbin:/usr/sbin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/opt/bin:/usr/armv7a-softfp-linux-gnueabi/gcc-bin/4.5.4:/root/bin:/root/bin > distcc[15991] exec on localhost: armv7a-softfp-linux-gnueabi-gcc -O2 -pipe > -march=armv7-a -mfpu=vfpv3-d16 -mfloat-abi=softfp test.c -o test > distcc[15991] (dcc_note_state) note state 5, file "(NULL)", host > "localhost" > distcc[15991] (dcc_spawn_child) forking to execute: > armv7a-softfp-linux-gnueabi-gcc -O2 -pipe -march=armv7-a -mfpu=vfpv3-d16 > -mfloat-abi=softfp test.c -o test > distcc[15991] (dcc_spawn_child) child started as pid15992 > distcc[15992] (dcc_increment_safeguard) setting safeguard: > _DISTCC_SAFEGUARD=2 > distcc[15991] (dcc_collect_child) cc child 15992 terminated with status 0 > distcc[15991] (dcc_collect_child) cc times: user 0.000000s, system > 0.000000s, 0 minflt, 0 majflt > distcc[15991] compile (null) on localhost completed ok > distcc[15991] elapsed compilation time 0.443756s > distcc[15991] (dcc_exit) exit: code 0; self: 0.000000 user 0.000000 sys; > children: 0.280000 user 0.120000 sys > distcc[15991] (dcc_cleanup_tempfiles_inner) deleted 0 temporary files > distcc[15990] (dcc_collect_child) cc child 15991 terminated with status 0 > distcc[15990] (dcc_collect_child) cc times: user 0.000000s, system > 0.000000s, 0 minflt, 0 majflt > distcc[15990] compile test.c on localhost completed ok > distcc[15990] (dcc_unlock) release lock fd4 > distcc[15990] elapsed compilation time 0.626678s > distcc[15990] (dcc_exit) exit: code 0; self: 0.000000 user 0.010000 sys; > children: 0.280000 user 0.120000 sys > distcc[15990] (dcc_cleanup_tempfiles_inner) deleted 0 temporary files > > > Using nmap on the beaglebone with > > nmap 192.168.178.21 -p 10-32000 gives > > prints: > > Starting Nmap 5.51 ( http://nmap.org ) at 2012-10-15 20:06 CEST > Nmap scan report for noname (192.168.178.21) > Host is up (0.00019s latency). > Not shown: 31987 closed ports > PORT STATE SERVICE > 22/tcp open ssh > 111/tcp open rpcbind > 2049/tcp open nfs > 3632/tcp open distccd > MAC Address: 00:15:F2:18:B0:20 (Asustek Computer) > > therefore distcc is listening on the PC but does not react in any way > -- eberything was compiled on the beaglebone (localhost). > > How can I convince the distccd on the PC to get its job done? > > Thank you very much in advance for any help! > > Best regards, > mcc > > > > __ > distcc mailing list http://distcc.samba.org/ > To unsubscribe or change options: > https://lists.samba.org/mailman/listinfo/distcc > -- Fergus Henderson <fer...@google.com> "Defend the user, exclude no one, and create magic." -- Eric Schmidt.
__ distcc mailing list http://distcc.samba.org/ To unsubscribe or change options: https://lists.samba.org/mailman/listinfo/distcc