x86, UV, BAU: Extend for more than 16 cpus per socket

Fix a hard-coded limit of a maximum of 16 cpu's per socket.

The UV Broadcast Assist Unit code initializes by scanning the
cpu topology of the system and assigning a master cpu for each
socket and UV hub. That scan had an assumption of a limit of 16
cpus per socket. With Westmere we are going over that limit.
The UV hub hardware will allow up to 32.

If the scan finds the system has gone over that limit it returns
an error and we print a warning and fall back to doing TLB
shootdowns without the BAU.

Signed-off-by: Cliff Wickman <[e-mail]>
Cc: <[e-mail]> # .37.x
LKML-Reference: <[e-mail]>
Signed-off-by: Ingo Molnar <[e-mail]>

arch/x86/include/asm/uv/uv_bau.h

···
  * BAU_SB_DESCRIPTOR_BASE register, set 1 is located at BASE + 512,
  * set 2 is at BASE + 2*512, set 3 at BASE + 3*512, and so on.
  *
- * We will use 31 sets, one for sending BAU messages from each of the 32
+ * We will use one set for sending BAU messages from each of the
  * cpu's on the uvhub.
  *
  * TLB shootdown will use the first of the 8 descriptors of each set.
  * Each of the descriptors is 64 bytes in size (8*64 = 512 bytes in a set).
  */
 
+#define MAX_CPUS_PER_UVHUB		64
+#define MAX_CPUS_PER_SOCKET		32
+#define UV_ADP_SIZE			64 /* hardware-provided max. */
+#define UV_CPUS_PER_ACT_STATUS		32 /* hardware-provided max. */
 #define UV_ITEMS_PER_DESCRIPTOR		8
 /* the 'throttle' to prevent the hardware stay-busy bug */
 #define MAX_BAU_CONCURRENT		3
-#define UV_CPUS_PER_ACT_STATUS		32
 #define UV_ACT_STATUS_MASK		0x3
 #define UV_ACT_STATUS_SIZE		2
-#define UV_ADP_SIZE			32
 #define UV_DISTRIBUTION_SIZE		256
 #define UV_SW_ACK_NPENDING		8
 #define UV_NET_ENDPOINT_INTD		0x38
···
  * number of destination side software ack resources
  */
 #define DEST_NUM_RESOURCES		8
-#define MAX_CPUS_PER_NODE		32
 /*
  * completion statuses for sending a TLB flush message
  */

arch/x86/platform/uv/tlb_uv.c

···
 
 	/*
 	 * each bau_desc is 64 bytes; there are 8 (UV_ITEMS_PER_DESCRIPTOR)
-	 * per cpu; and up to 32 (UV_ADP_SIZE) cpu's per uvhub
+	 * per cpu; and one per cpu on the uvhub (UV_ADP_SIZE)
 	 */
 	bau_desc = kmalloc_node(sizeof(struct bau_desc) * UV_ADP_SIZE
 				* UV_ITEMS_PER_DESCRIPTOR, GFP_KERNEL, node);
···
 /*
  * initialize the bau_control structure for each cpu
  */
-static void __init uv_init_per_cpu(int nuvhubs)
+static int __init uv_init_per_cpu(int nuvhubs)
 {
 	int i;
 	int cpu;
···
 	struct bau_control *smaster = NULL;
 	struct socket_desc {
 		short num_cpus;
-		short cpu_number[16];
+		short cpu_number[MAX_CPUS_PER_SOCKET];
 	};
 	struct uvhub_desc {
 		unsigned short socket_mask;
···
 		sdp = &bdp->socket[socket];
 		sdp->cpu_number[sdp->num_cpus] = cpu;
 		sdp->num_cpus++;
+		if (sdp->num_cpus > MAX_CPUS_PER_SOCKET) {
+			printk(KERN_EMERG "%d cpus per socket invalid\n", sdp->num_cpus);
+			return 1;
+		}
 	}
 	for (uvhub = 0; uvhub < nuvhubs; uvhub++) {
 		if (!(*(uvhub_mask + (uvhub/8)) & (1 << (uvhub%8))))
···
 				bcp->uvhub_master = hmaster;
 				bcp->uvhub_cpu = uv_cpu_hub_info(cpu)->
 						blade_processor_id;
+				if (bcp->uvhub_cpu >= MAX_CPUS_PER_UVHUB) {
+					printk(KERN_EMERG
+						"%d cpus per uvhub invalid\n",
+						bcp->uvhub_cpu);
+					return 1;
+				}
 			}
 nextsocket:
 			socket++;
···
 		bcp->congested_reps = congested_reps;
 		bcp->congested_period = congested_period;
 	}
+	return 0;
 }
 
 /*
···
 	spin_lock_init(&disable_lock);
 	congested_cycles = microsec_2_cycles(congested_response_us);
 
-	uv_init_per_cpu(nuvhubs);
+	if (uv_init_per_cpu(nuvhubs)) {
+		nobau = 1;
+		return 0;
+	}
 
 	uv_partition_base_pnode = 0x7fffffff;
 	for (uvhub = 0; uvhub < nuvhubs; uvhub++)