From mboxrd@z Thu Jan 1 00:00:00 1970 From: Brent Casavant Date: Wed, 25 Jan 2006 19:15:49 +0000 Subject: RE: [PATCH] SN2 user-MMIO CPU migration Message-Id: <20060125131008.I16092@chenjesu.americas.sgi.com> List-Id: References: <20060118163305.Y42462@chenjesu.americas.sgi.com> In-Reply-To: <20060118163305.Y42462@chenjesu.americas.sgi.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-ia64@vger.kernel.org On Wed, 25 Jan 2006, Chen, Kenneth W wrote: > I suppose, I don't have strong opinion where it is being called. > Adding to ia64_load_extra(), you might need #ifdef CONFIG_SMP, of > course you can just sneak in a plain version unless people yell at > you for slowing their UP performance (I won't yell at you for UP). Hmm. I just thought of a good reason. Adding it to IA64_HAS_EXTRA_STATE will cause ia64_{load,save}_extra to be invoked with every single context switch on SN2, which I don't think we want to do. Better to leave it a separate test. We may want to revisit that in the future if/when IA64_THREAD_MIGRATION indicates something finer-grained than "Is/isn't SN2", but for now it's the better solution. > > __switch_task() (actually, after ia64_switch_to()) I would receive > > kernel panics during boot (the migration threads would die from an > > invalid access, swapper shortly thereafter, and finally a "soft > lockup" > > on swapper). Was I perhaps missing something? > > I can't immediately see why it won't work on sn2. It works for me > on a Intel tiger ia64 machine. OK, I'll dig deeper into it. I suspect ia64_switch_to() is changing something important underneath us (e.g. a value saved on the stack, which would tend to change when we switch stacks). Brent -- Brent Casavant All music is folk music. I ain't bcasavan@sgi.com never heard a horse sing a song. Silicon Graphics, Inc. -- Louis Armstrong