Hi Roland, On Fri, 2006-02-03 at 11:58, Roland Dreier wrote: > Linus just released 2.6.16-rc2. If you have a chance, please test the > IB drivers in the stock kernel (ie don't replace drivers/infiniband > with a svn tree) and report any issues you see. Also, please make > sure that any important fixes that you would like to see in 2.6.16 are > upstream -- if not, let me know so we can get them merged.
Perhaps not critical for 2.6.16, but should be pushed upstream are the following SMI changes: r5045 | halr | 2006-01-17 11:44:14 -0500 (Tue, 17 Jan 2006) | 6 lines Changed paths: M /gen2/trunk/src/linux-kernel/infiniband/core/mad.c Simplified patch to properly handle directed route SMP with a beginning or ending LID routed part. Signed-off-by: Ralph Campbell <[EMAIL PROTECTED]> Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]> ------------------------------------------------------------------------ r4984 | halr | 2006-01-14 13:08:55 -0500 (Sat, 14 Jan 2006) | 15 lines Changed paths: M /gen2/trunk/src/linux-kernel/infiniband/core/agent.c M /gen2/trunk/src/linux-kernel/infiniband/core/mad.c M /gen2/trunk/src/linux-kernel/infiniband/core/smi.h Further simplification of SMI by eliminating smi_check_local_dr_smp The call to ib_get_agent_port() shouldn't be possible to fail when smi_check_local_dr_smp() is called from ib_mad_recv_done_handler(). When it is called from handle_outgoing_dr_smp(), the device and port_num come from mad_agent_priv so I assume the call to ib_get_agent_port() shouldn't fail either. In either case, smi_check_local_smp() only uses the mad_agent pointer to check that mad_agent->device->process_mad is not NULL. The device pointer would have to be the same as the one passed to smi_check_local_dr_smp() since that pointer is used later instead of the one checked in smi_check_local_smp(). Patch supplied by Ralph Campbell <[EMAIL PROTECTED]> Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]> ------------------------------------------------------------------------ r4983 | halr | 2006-01-14 12:42:08 -0500 (Sat, 14 Jan 2006) | 10 lines Changed paths: M /gen2/trunk/src/linux-kernel/infiniband/core/agent.c Remove redundant check from agent.c::smi_check_local_dr_smp smi_check_local_dr_smp() is called only from two places in core/mad.c It returns 0 or 1. In smi_check_local_dr_smp(), it checks for a directed route SMP but this function is only called when the SMP is a directed route so this is a NOP. Patch supplied by Ralph Campbell <[EMAIL PROTECTED]> Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]> Thanks. -- Hal > Thanks, > Roland > _______________________________________________ > openib-general mailing list > [email protected] > http://openib.org/mailman/listinfo/openib-general > > To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general _______________________________________________ openib-general mailing list [email protected] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
