From: Frank P. <pi...@we...> - 2006-02-26 14:09:46
|
Hi, since 1.2.9 there seems to be ipset as new feature, but it breaks my compile. I don't find a config switch for that. make[1]: Entering directory `/build/tmp/ipset-2.2.8' cc -O2 -fno-stack-protector -Wall -Wunused -I/build/tmp/linux-2.4.32/include -I. -DIPSET_VERSION=\"2.2.8\" -DIPSE T_LIB_DIR=\"/usr/lib/ipset\" -c -o ipset.o ipset.c In file included from ipset.c:26: ipset.h:27:41: linux/netfilter_ipv4/ip_set.h: No such file or directory In file included from ipset.c:26: ipset.h:79: error: `IP_SET_MAXNAMELEN' undeclared here (not in a function) ..... But the header seems to be there: ./tmp/patch-o-matic-ng-20060211/patchlets/set/linux/include/linux/netfilter_ipv4/ip_set.h in menuconfig I don't have selected netfilter patches, is this a precondition? Regards Frank |
From: Serge L. <fi...@in...> - 2006-02-27 08:57:54
|
Dear Frank, Sunday, February 26, 2006, 5:15:52 PM, you wrote: > Hi, > since 1.2.9 there seems to be ipset as new feature, but it breaks my compile. > I don't find a config switch for that. ... > in menuconfig I don't have selected netfilter patches, is this a precondition? Yes, the ipset requires base repository of p-o-m. Thank you for the bug-report. I'll fix it by checking of p-o-m settings. -- Best regards, Serge mailto:fi...@in... |
From: Frank P. <pi...@we...> - 2006-02-27 19:05:41
|
On Monday 27 February 2006 09:57, Serge Leschinsky wrote: > Dear Frank, Dear Serge, thanks for your fast fix, BUT .... > > > in menuconfig I don't have selected netfilter patches, is this a > > precondition? > > Yes, the ipset requires base repository of p-o-m. Thank you for the > bug-report. I'll fix it by checking of p-o-m settings. it don't work :-( You forgot the "if" in _both_ of the if [ "$CONFIG_NF_BASE" = "y" ]; then lines in scripts/ipset (produces a syntax error). This one was easy to fix, now it stops in "util-linux-2.12r/misc-utils" Will see ... Regards Frank |
From: Serge L. <fi...@in...> - 2006-02-28 06:45:56
|
Dear Frank, Monday, February 27, 2006, 10:10:55 PM, you wrote: >> Yes, the ipset requires base repository of p-o-m. Thank you for the >> bug-report. I'll fix it by checking of p-o-m settings. > it don't work :-( > You forgot the "if" in _both_ of the > if [ "$CONFIG_NF_BASE" = "y" ]; then > lines in scripts/ipset (produces a syntax error). Copy-paste :-( Fixed once again. > This one was easy to fix, now it stops in "util-linux-2.12r/misc-utils" > Will see ... OK. I'll check it -- Best regards, Serge mailto:fi...@in... |
From: Frank P. <pi...@we...> - 2006-02-27 21:28:33
|
On Monday 27 February 2006 20:10, Frank Pieczynski wrote: > now it stops in "util-linux-2.12r/misc-utils" > Will see ... seems to be also a issue with parallel build. Some object files have a length of zero, the linker complains about missing main. After clean and setting the number of jobs in menuconfig to 1 it runs through. Regards Frank |
From: Serge L. <fi...@in...> - 2006-02-28 07:20:24
|
Dear Frank, Tuesday, February 28, 2006, 12:35:01 AM, you wrote: >> now it stops in "util-linux-2.12r/misc-utils" >> Will see ... > seems to be also a issue with parallel build. Some object files have a length > of zero, the linker complains about missing main. > After clean and setting the number of jobs in menuconfig to 1 it runs through. Could you please do the following changes for scripts/util-linux and try to do a parallel build again? ( I just delete $PMAKE from the "make" line - I think it should be equivalent to setting the number of jobs in menuconfig to 1 but for this script only) > case $1 in > build ) ... > replace_str defines.h "#define ENABLE_NLS" "" > if [ "$CONFIG_GCC_ET_DYN" = "y" ]; then > # make $PMAKE HAVE_SLN=yes ADD_RAW=yes DISABLE_NLS=yes LDFLAGS="-pie" || exit 1 > make HAVE_PAM=yes HAVE_SLN=yes ADD_RAW=yes DISABLE_NLS=yes LDFLAGS="-pie" || exit 1 > else > # make $PMAKE HAVE_SLN=yes ADD_RAW=yes DISABLE_NLS=yes || exit 1 > make HAVE_PAM=yes HAVE_SLN=yes ADD_RAW=yes DISABLE_NLS=yes || exit 1 > fi > strip_debug > ;; -- Best regards, Serge mailto:fi...@in... |
From: Frank P. <pi...@we...> - 2006-03-05 16:59:10
|
On Tuesday 28 February 2006 08:20, Serge Leschinsky wrote: Dear Serge, > Tuesday, February 28, 2006, 12:35:01 AM, you wrote: > >> now it stops in "util-linux-2.12r/misc-utils" > >> Will see ... > > > > seems to be also a issue with parallel build. Some object files have a > > length of zero, the linker complains about missing main. > > After clean and setting the number of jobs in menuconfig to 1 it runs > > through. > > Could you please do the following changes for scripts/util-linux and > try to do a parallel build again? > ( I just delete $PMAKE from the "make" line - I think it should be > equivalent to setting the number of jobs in menuconfig to 1 but for > this script only) > > > case $1 in > > build ) > > ... > > > replace_str defines.h "#define ENABLE_NLS" "" > > if [ "$CONFIG_GCC_ET_DYN" = "y" ]; then > > # make $PMAKE HAVE_SLN=yes ADD_RAW=yes DISABLE_NLS=yes > > LDFLAGS="-pie" || exit 1 make HAVE_PAM=yes HAVE_SLN=yes ADD_RAW=yes > > DISABLE_NLS=yes LDFLAGS="-pie" || exit 1 else > > # make $PMAKE HAVE_SLN=yes ADD_RAW=yes DISABLE_NLS=yes > > || exit 1 make HAVE_PAM=yes HAVE_SLN=yes ADD_RAW=yes DISABLE_NLS=yes || > > exit 1 fi > > strip_debug > > ;; Sorry, I tried a lot of times, but I cannot reproduce the problem, so I cannot test really your fix. May be timing, disk activities ... I have also tried a complete fresh compile today after 'make mrproper'. It stopped once in ncurses, but runs through util-linux without problems. Thanks, Frank |