Ed, Thanks! Suggest to create Feature Request ticket in SF OpenHPI tracker about FreeBSD porting.
https://sourceforge.net/tracker/?group_id=71730&atid=532254 4) We have .cpp file but not .s. There is no weak references in it. Seems .s was a temporary product of gcc/g++. I'll look into other issues. Anton Pak On Tue, 15 Feb 2011 05:51:57 +0300, Ed Maste <[email protected]> wrote: > I'm doing an initial investigation into porting OpenHPI to FreeBSD, and > have a couple of observations so far. > > 1) GNU configure - one of the tests is using == for string comparison, > which works if /bin/sh is bash, but not if it's FreeBSD's sh. (I > suspect Debian and Ubuntu users will encounter this too, with /bin/sh > being dash.) > > The fix is: > > $ diff -purw acinclude.m4.orig acinclude.m4 > --- acinclude.m4.orig 2011-02-14 21:15:51.000000000 -0500 > +++ acinclude.m4 2011-02-14 21:16:10.000000000 -0500 > @@ -137,7 +137,7 @@ AC_DEFUN([OH_CHECK_GCC], > } \ > }'` \ > - if test "$GCCVERSIONOK" == "OK"; then > + if test "$GCCVERSIONOK" = "OK"; then > AC_MSG_RESULT(yes) > else > OH_CHECK_FAIL(gcc >= $1.$2.$3 is required to build OpenHPI) > @@ -200,7 +200,7 @@ AC_DEFUN([OH_CHECK_OPENIPMI], > }'` > fi > - if test "$OPENIPMI_VERSION" == "OK"; then > + if test "$OPENIPMI_VERSION" = "OK"; then > have_openipmi=yes > AC_MSG_RESULT(yes) > else > > 2) DST_* conflict > > In plugins/snmp_bc/snmp_bc_time.h there's an enum with members DST_NONE, > DST_USA, etc. These conflict with #DEFINEs of DST_NONE, DST_USA and the > like in FreeBSD's <sys/time.h> include. The actual enums don't seem to > be used in the source; I'm not yet familiar enough with OpenHPI to know > how they're supposed to be used. > > 3) PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP not defined. > > Used in plugins/ipmidirect/thread.cpp . With the _NP suffix implying > nonportable it's not surprising that this doesn't work. > > 4) Unknown assembly pseudo-op .weakref > > Used in plugins/ipmidirect/ipmi_mc_vendor_sun.s . It looks like the > stock binutils used by FreeBSD is just too old, so the port will just > have to use an explicit newer version. > > Right now this is just an initial investigation - if I continue on and > make progress I'll report back. > > Regards, > Ed > > ------------------------------------------------------------------------------ > The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: > Pinpoint memory and threading errors before they happen. > Find and fix more than 250 security defects in the development cycle. > Locate bottlenecks in serial and parallel code that limit performance. > http://p.sf.net/sfu/intel-dev2devfeb > _______________________________________________ > Openhpi-devel mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/openhpi-devel ------------------------------------------------------------------------------ The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb _______________________________________________ Openhpi-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/openhpi-devel
