On Sun, 2010-03-07 at 15:08 -0800, Hollis Blanchard wrote: > On Fri, Mar 5, 2010 at 12:43 PM, Dave Kleikamp > <sha...@linux.vnet.ibm.com> wrote: > > > > powerpc/476: Add isync after loading mmu and debug spr's > > > > From: Dave Kleikamp <sha...@linux.vnet.ibm.com> > > > > 476 requires an isync after loading MMU and debug related SPR's. Some of > > these are in performance-critical paths and may need to be optimized, but > > initially, we're playing it safe. > > Why is there sometimes #ifdef CONFIG_PPC_47x, sometimes not, and never > the cputable infrastructure which was intended for this sort of thing?
The places without an ifdef are either in cpu initialization code, where the cost of an isync is insignificant or, in one case, in 47x-specific code. I was having problems trying to get the cputable infrastructure compiling in inline assembly in the *.c files, but I could have used that in some places. Currently, one can't build with CONFIG_PPC_47x for a non-47x cpu, and we want to re-evaluate whether we can get by without these isyncs. So basically, I plan to clean this up somehow by the time we have a working unified 44x/47x binary kernel. Thanks, Shaggy -- David Kleikamp IBM Linux Technology Center _______________________________________________ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
