<br><br>
<div class="gmail_quote">On Thu, Sep 17, 2009 at 5:40 PM, Ira Weiny <span dir="ltr"><<a href="mailto:weiny2@llnl.gov">weiny2@llnl.gov</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">
<div>
<div></div>
<div class="h5">On Thu, 17 Sep 2009 17:41:30 -0400<br>Hal Rosenstock <<a href="mailto:hal.rosenstock@gmail.com">hal.rosenstock@gmail.com</a>> wrote:<br><br>> On Thu, Sep 17, 2009 at 4:20 PM, Ira Weiny <<a href="mailto:weiny2@llnl.gov">weiny2@llnl.gov</a>> wrote:<br>
><br>> > On Thu, 17 Sep 2009 10:35:39 -0700<br>> > "Sean Hefty" <<a href="mailto:sean.hefty@intel.com">sean.hefty@intel.com</a>> wrote:<br>> ><br>> > > >> #define IB_PATH_RECORD_REVERSIBLE 0x80<br>
> > > >><br>> > > >> struct ib_path_record<br>> > > >> {<br>> > > >> uint64_t service_id;<br>> > > >> union ibv_gid dgid;<br>> > > >> union ibv_gid sgid;<br>
> > > >> uint16_t dlid;<br>> > > >> uint16_t slid;<br>> > > >> uint32_t flowlabel_hoplimit; /* resv-31:28 flow label-27:8<br>> > hop<br>> > > >limit-7:0*/<br>
> > > >> uint8_t tclass;<br>> > > >> uint8_t reversible_numpath; /* reversible-7:7 num path-6:0<br>> > */<br>> > > >> uint16_t pkey;<br>> > > >> uint16_t qosclass_sl; /* qos class-15:4 sl-3:0 */<br>
> > > >> uint8_t mtu; /* mtu selector-7:6 mtu-5:0 */<br>> > > >> uint8_t rate; /* rate selector-7:6 rate-5:0<br>> > */<br>> > > >> uint8_t packetlifetime; /* lifetime selector-7:6<br>
> > > lifetime-5:0<br>> > > >*/<br>> > > >> uint8_t preference;<br>> > > >> uint8_t reserved[6];<br>> > > >> };<br>> > > ><br>
> > > >I would prefer to use the structures already defined in ib_types.h... I<br>> > > >understand your not wanting to make ACM dependant on the OpenSM packages<br>> > so is<br>> > > >it time to move ib_types.h out of the OpenSM tree and somewhere more<br>
> > generic?<br>> > > >Perhaps libibumad? This also applies to ib_sa_mad in your 5th patch.<br>> > > ><br>> > > >OTOH, ib_types.h is a 10K line file with multiple long (>10 lines)<br>
> > inlined<br>> > > >functions. Perhaps it deserves it's own library?<br>> > ><br>> > > Defining some of these types in libibumad isn't a bad idea. Although,<br>> > WinOF<br>
> > > actually has 2 copies of ib_types.h (that differ...) I find using<br>> > ib_types.h<br>> > > painful given its size; separate header files may help.<br>> ><br>> > Yes I was thinking multiple headers. There seems like there is already<br>
> > some precedent in ib_cm_types.h (although that entire file seems to be<br>> > enclosed in a #ifndef WIN32 clause? So am I wrong on this?)<br>> ><br>> > In the end I would like to make ib_types.h just list the specific headers.<br>
> ><br>> > Sasha, would you be willing to accept such a patch? First move ib_types.h<br>> > to umad<br>><br>><br>> I'm not sure this is a good idea. ibutils (ibis and ibmgtsim) wants<br>> ib_types.h but does not want libibumad.<br>
</div></div></blockquote>
<div> </div>
<div>I miswrote about ibis as it uses osm_vendor layer so it can use libibumad but there are other vendor layers other than osm_vendor_ibumad in use.</div>
<div> </div>
<div>There are other combinations where umad isn't used (even Windows is not fully moved over still).</div>
<div> </div>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">
<div>
<div class="h5"><br></div></div>Would a separate library be a better solution then? </blockquote>
<div> </div>
<div>Maybe but what aside from the header would be in the library ?</div>
<div> </div>
<div>-- Hal</div>
<div> </div>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid"> I would prefer that as well.<br><font color="#888888"><br>Ira<br></font>
<div class="im"><br>><br>> -- Hal<br>><br>><br>><br>> > and then move the long inline functions into the lib and separate out the<br>> > remaining header.<br>> ><br>> > Or would you prefer a new library? I think there is enough code there but<br>
> > I leave it up to you.<br>> ><br>> > Ira<br>> ><br>> > ><br>> > > - Sean<br>> > ><br>> ><br>> ><br>> > --<br>> > Ira Weiny<br>> > Math Programmer/Computer Scientist<br>
> > Lawrence Livermore National Lab<br>> > 925-423-8008<br>> > <a href="mailto:weiny2@llnl.gov">weiny2@llnl.gov</a><br>> > _______________________________________________<br>> > ofw mailing list<br>
> > <a href="mailto:ofw@lists.openfabrics.org">ofw@lists.openfabrics.org</a><br>> > http://*<a href="http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw" target="_blank">lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw</a><br>
> ><br>><br><br><br></div>--<br>
<div>
<div></div>
<div class="h5">Ira Weiny<br>Math Programmer/Computer Scientist<br>Lawrence Livermore National Lab<br>925-423-8008<br><a href="mailto:weiny2@llnl.gov">weiny2@llnl.gov</a><br></div></div></blockquote></div><br>