x86/uv: Update the UV3 TLB shootdown logic

Update of TLB shootdown code for UV3.

Kernel function native_flush_tlb_others() calls
uv_flush_tlb_others() on UV to invalidate tlb page definitions
on remote cpus. The UV systems have a hardware 'broadcast assist
unit' which can be used to broadcast shootdown messages to all
cpu's of selected nodes.

The behavior of the BAU has changed only slightly with UV3:

  - UV3 is recognized with is_uv3_hub().
  - UV2 functions and structures (uv2_xxx) are in most cases
    simply renamed to uv2_3_xxx.
  - Some UV2 error workarounds are not needed for UV3.
    (see uv_bau_message_interrupt and enable_timeouts)

Signed-off-by: Cliff Wickman <cpw@sgi.com>
Link: http://lkml.kernel.org/r/E1WkgWh-0001yJ-3K@eag09.americas.sgi.com
[ Removed a few linebreak uglies. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>

2 files changed, 49 insertions, 39 deletions

diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h
index 0b46ef261c77..2d60a7813dfe 100644
--- a/arch/x86/include/asm/uv/uv_bau.h
+++ b/arch/x86/include/asm/uv/uv_bau.h
@@ -73,6 +73,7 @@
 #define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD (is_uv1_hub() ?                 \
                 UV1_INTD_SOFT_ACK_TIMEOUT_PERIOD :                      \
                 UV2_INTD_SOFT_ACK_TIMEOUT_PERIOD)
+/* assuming UV3 is the same */
 
 #define BAU_MISC_CONTROL_MULT_MASK      3
 
@@ -93,6 +94,8 @@
 #define SOFTACK_MSHIFT UVH_LB_BAU_MISC_CONTROL_ENABLE_INTD_SOFT_ACK_MODE_SHFT
 #define SOFTACK_PSHIFT UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHFT
 #define SOFTACK_TIMEOUT_PERIOD UV_INTD_SOFT_ACK_TIMEOUT_PERIOD
+#define PREFETCH_HINT_SHFT UV3H_LB_BAU_MISC_CONTROL_ENABLE_INTD_PREFETCH_HINT_SHFT
+#define SB_STATUS_SHFT UV3H_LB_BAU_MISC_CONTROL_ENABLE_EXTENDED_SB_STATUS_SHFT
 #define write_gmmr      uv_write_global_mmr64
 #define write_lmmr      uv_write_local_mmr
 #define read_lmmr       uv_read_local_mmr
@@ -322,8 +325,9 @@ struct uv1_bau_msg_header {
 /*
  * UV2 Message header:  16 bytes (128 bits) (bytes 0x30-0x3f of descriptor)
  * see figure 9-2 of harp_sys.pdf
+ * assuming UV3 is the same
  */
-struct uv2_bau_msg_header {
+struct uv2_3_bau_msg_header {
         unsigned int    base_dest_nasid:15;     /* nasid of the first bit */
         /* bits 14:0 */                         /* in uvhub map */
         unsigned int    dest_subnodeid:5;       /* must be 0x10, for the LB */
@@ -395,7 +399,7 @@ struct bau_desc {
          */
         union bau_msg_header {
                 struct uv1_bau_msg_header       uv1_hdr;
-                struct uv2_bau_msg_header       uv2_hdr;
+                struct uv2_3_bau_msg_header     uv2_3_hdr;
         } header;
 
         struct bau_msg_payload                  payload;
@@ -631,11 +635,6 @@ struct bau_control {
         struct hub_and_pnode    *thp;
 };
 
-static inline unsigned long read_mmr_uv2_status(void)
-{
-        return read_lmmr(UV2H_LB_BAU_SB_ACTIVATION_STATUS_2);
-}
-
 static inline void write_mmr_data_broadcast(int pnode, unsigned long mmr_image)
 {
         write_gmmr(pnode, UVH_BAU_DATA_BROADCAST, mmr_image);
@@ -760,7 +759,11 @@ static inline int atomic_read_short(const struct atomic_short *v)
  */
 static inline int atom_asr(short i, struct atomic_short *v)
 {
-        return i + xadd(&v->counter, i);
+        short __i = i;
+        asm volatile(LOCK_PREFIX "xaddw %0, %1"
+                        : "+r" (i), "+m" (v->counter)
+                        : : "memory");
+        return i + __i;
 }
 
 /*
diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c
index dfe605ac1bcd..ed161c6e278b 100644
--- a/arch/x86/platform/uv/tlb_uv.c
+++ b/arch/x86/platform/uv/tlb_uv.c
@@ -1,7 +1,7 @@
 /*
  *      SGI UltraViolet TLB flush routines.
  *
- *      (c) 2008-2012 Cliff Wickman <cpw@sgi.com>, SGI.
+ *      (c) 2008-2014 Cliff Wickman <cpw@sgi.com>, SGI.
  *
  *      This code is released under the GNU General Public License version 2 or
  *      later.
@@ -451,7 +451,7 @@ static inline unsigned long long cycles_2_ns(unsigned long long cyc)
 
 /*
  * The reverse of the above; converts a duration in ns to a duration in cycles.
- */ 
+ */
 static inline unsigned long long ns_2_cycles(unsigned long long ns)
 {
         struct cyc2ns_data *data = cyc2ns_read_begin();
@@ -563,7 +563,7 @@ static int uv1_wait_completion(struct bau_desc *bau_desc,
  * UV2 could have an extra bit of status in the ACTIVATION_STATUS_2 register.
  * But not currently used.
  */
-static unsigned long uv2_read_status(unsigned long offset, int rshft, int desc)
+static unsigned long uv2_3_read_status(unsigned long offset, int rshft, int desc)
 {
         unsigned long descriptor_status;
 
@@ -606,7 +606,7 @@ int handle_uv2_busy(struct bau_control *bcp)
         return FLUSH_GIVEUP;
 }
 
-static int uv2_wait_completion(struct bau_desc *bau_desc,
+static int uv2_3_wait_completion(struct bau_desc *bau_desc,
                                 unsigned long mmr_offset, int right_shift,
                                 struct bau_control *bcp, long try)
 {
@@ -616,7 +616,7 @@ static int uv2_wait_completion(struct bau_desc *bau_desc,
         long busy_reps = 0;
         struct ptc_stats *stat = bcp->statp;
 
-        descriptor_stat = uv2_read_status(mmr_offset, right_shift, desc);
+        descriptor_stat = uv2_3_read_status(mmr_offset, right_shift, desc);
 
         /* spin on the status MMR, waiting for it to go idle */
         while (descriptor_stat != UV2H_DESC_IDLE) {
@@ -658,8 +658,7 @@ static int uv2_wait_completion(struct bau_desc *bau_desc,
                                 /* not to hammer on the clock */
                                 busy_reps = 0;
                                 ttm = get_cycles();
-                                if ((ttm - bcp->send_message) >
-                                                bcp->timeout_interval)
+                                if ((ttm - bcp->send_message) > bcp->timeout_interval)
                                         return handle_uv2_busy(bcp);
                         }
                         /*
@@ -667,8 +666,7 @@ static int uv2_wait_completion(struct bau_desc *bau_desc,
                          */
                         cpu_relax();
                 }
-                descriptor_stat = uv2_read_status(mmr_offset, right_shift,
-                                                                        desc);
+                descriptor_stat = uv2_3_read_status(mmr_offset, right_shift, desc);
         }
         bcp->conseccompletes++;
         return FLUSH_COMPLETE;
@@ -679,8 +677,7 @@ static int uv2_wait_completion(struct bau_desc *bau_desc,
  * which register to read and position in that register based on cpu in
  * current hub.
  */
-static int wait_completion(struct bau_desc *bau_desc,
-                                struct bau_control *bcp, long try)
+static int wait_completion(struct bau_desc *bau_desc, struct bau_control *bcp, long try)
 {
         int right_shift;
         unsigned long mmr_offset;
@@ -695,11 +692,9 @@ static int wait_completion(struct bau_desc *bau_desc,
         }
 
         if (bcp->uvhub_version == 1)
-                return uv1_wait_completion(bau_desc, mmr_offset, right_shift,
-                                                                bcp, try);
+                return uv1_wait_completion(bau_desc, mmr_offset, right_shift, bcp, try);
         else
-                return uv2_wait_completion(bau_desc, mmr_offset, right_shift,
+                return uv2_3_wait_completion(bau_desc