Re: CVS commit: src/regress/sys/net

2011-03-03 Thread David Laight
On Thu, Mar 03, 2011 at 12:27:19AM +, David Holland wrote: On Tue, Mar 01, 2011 at 08:02:40AM +, David Laight wrote: autogen requires ed(1), which is not provided by our toolchain. (In other words, this causes a odd build failure on some Linux build hosts.) I've not

Re: CVS commit: src/regress/sys/net

2011-03-03 Thread David Holland
On Thu, Mar 03, 2011 at 08:09:49AM +, David Laight wrote: +awk $1 ' +state == 0 /^ether_aton_r/ { print prev; state = 1; } +state == 1 { print; } +state == 1 /^}$/ { state = 2; } +{ prev = $0; } +' $2 Looks like that would be even simpler in sed :-)

Re: CVS commit: src/regress/sys/net

2011-03-02 Thread David Holland
On Tue, Mar 01, 2011 at 08:02:40AM +, David Laight wrote: autogen requires ed(1), which is not provided by our toolchain. (In other words, this causes a odd build failure on some Linux build hosts.) I've not looked, but it is probably not too hard to change the script to use a

Re: CVS commit: src/regress/sys/net

2011-03-02 Thread David Holland
On Thu, Mar 03, 2011 at 12:27:19AM +, David Holland wrote: This should do the trick, modulo feeding in ${TOOL_AWK} properly: (and if anyone's wondering why I haven't just committed it, it's because once again I've forgotten how to persuade atf to work...) -- David A. Holland

Re: CVS commit: src/regress/sys/net

2011-02-28 Thread David Laight
On Mon, Feb 28, 2011 at 10:08:07PM +, Jonathan A. Kollasch wrote: autogen requires ed(1), which is not provided by our toolchain. (In other words, this causes a odd build failure on some Linux build hosts.) I've not looked, but it is probably not too hard to change the script to use a