[IA64] speed up syscall path a bit more

Recently I noticed that clearing ar.ssd/ar.csd right before srlz.d is
causing significant stalling in the syscall path.  The patch below
fixes that by moving the register-writes after srlz.d.  On a Madison,
this drops break-based getpid() from 241 to 226 cycles (-15 cycles).

Signed-off-by: David Mosberger-Tang <davidm@hpl.hp.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

1 files changed, 6 insertions, 6 deletions

diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S
index 0272c010a3ba..73e23dafe8e9 100644
--- a/arch/ia64/kernel/entry.S
+++ b/arch/ia64/kernel/entry.S
@@ -728,12 +728,8 @@ ENTRY(ia64_leave_syscall)
         mov f8=f0               // clear f8
         ;;
         ld8 r30=[r2],16         // M0|1 load cr.ifs
-        mov.m ar.ssd=r0         // M2 clear ar.ssd
-        cmp.eq p9,p0=r0,r0      // set p9 to indicate that we should restore cr.ifs
-        ;;
         ld8 r25=[r3],16         // M0|1 load ar.unat
-        mov.m ar.csd=r0         // M2 clear ar.csd
-        mov r22=r0              // clear r22
+        cmp.eq p9,p0=r0,r0      // set p9 to indicate that we should restore cr.ifs
         ;;
         ld8 r26=[r2],PT(B0)-PT(AR_PFS)  // M0|1 load ar.pfs
 (pKStk) mov r22=psr             // M2 read PSR now that interrupts are disabled
@@ -756,11 +752,15 @@ ENTRY(ia64_leave_syscall)
         mov f7=f0               // clear f7
         ;;
         ld8.fill r12=[r2]       // restore r12 (sp)
+        mov.m ar.ssd=r0         // M2 clear ar.ssd
+        mov r22=r0              // clear r22
+
         ld8.fill r15=[r3]       // restore r15
+(pUStk) st1 [r14]=r17
         addl r3=THIS_CPU(ia64_phys_stacked_size_p8),r0
         ;;
 (pUStk) ld4 r3=[r3]             // r3 = cpu_data->phys_stacked_size_p8
-(pUStk) st1 [r14]=r17
+        mov.m ar.csd=r0         // M2 clear ar.csd
         mov b6=r18              // I0  restore b6
         ;;
         mov r14=r0              // clear r14