Hi Tony: When returning to the kernel and not switching rbs, avoid fetching local_cpu_data. More could possibly be optimized here but limited registers are available and the number of bundles might increase. I didn't pursue any further than predicating two instructions which eliminates a memory fetch. thanks, bob Signed-off-by: Bob Picco <robert.picco@hp.com> diff -ruNp -X /home/picco/losl/dontdiff linux-2.6.10-orig/arch/ia64/kernel/entry.S linux-2.6.10/arch/ia64/kernel/entry.S --- linux-2.6.10-orig/arch/ia64/kernel/entry.S 2005-01-04 19:06:14.000000000 -0500 +++ linux-2.6.10/arch/ia64/kernel/entry.S 2005-01-08 15:47:45.000000000 -0500 @@ -925,9 +925,9 @@ GLOBAL_ENTRY(ia64_leave_kernel) shr.u r18=r19,16 // get byte size of existing "dirty" partition ;; mov r16=ar.bsp // get existing backing store pointer - addl r17=THIS_CPU(ia64_phys_stacked_size_p8),r0 +(pUStk) addl r17=THIS_CPU(ia64_phys_stacked_size_p8),r0 ;; - ld4 r17=[r17] // r17 = cpu_data->phys_stacked_size_p8 +(pUStk) ld4 r17=[r17] // r17 = cpu_data->phys_stacked_size_p8 (pKStk) br.cond.dpnt skip_rbs_switch /* - To unsubscribe from this list: send the line "unsubscribe linux-ia64" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.htmlReceived on Tue Jan 11 20:39:28 2005
This archive was generated by hypermail 2.1.8 : 2005-08-02 09:20:34 EST