On 25.05.2011 21:38, minf...@apache.org wrote: > Author: minfrin > Date: Wed May 25 19:38:23 2011 > New Revision: 1127648 > > URL: http://svn.apache.org/viewvc?rev=1127648&view=rev > Log: > apr_crypto: Make sure we can find the nss headers on RHEL5. > > Modified: > apr/apr/trunk/build/crypto.m4 > > Modified: apr/apr/trunk/build/crypto.m4 > URL: > http://svn.apache.org/viewvc/apr/apr/trunk/build/crypto.m4?rev=1127648&r1=1127647&r2=1127648&view=diff > ============================================================================== > --- apr/apr/trunk/build/crypto.m4 (original) > +++ apr/apr/trunk/build/crypto.m4 Wed May 25 19:38:23 2011 > @@ -163,7 +163,7 @@ AC_DEFUN([APU_CHECK_CRYPTO_NSS], [ > apu_have_nss=0 > elif test "x$withval" != "x"; then > > - nss_CPPFLAGS="-I$withval/include -I$withval/../public" > + nss_CPPFLAGS="-I$withval/include/nss -I$withval/include/nss3 > -I$withval/include/nspr -I$withval/include/nspr4 -I$withval/include > -I$withval/../public" > nss_LDFLAGS="-L$withval/lib " > > APR_ADDTO(CPPFLAGS, [$nss_CPPFLAGS])
Would it make sense to detect the correct path? Having many possibly conflicting include paths could be difficult to debug. E.g. for one version the base rpm is installed, for another one the dev with headers and we cannot easily find out, which headers were chosen. In addition: shouldn't the most recent version come first? Regards, Rainer
