Pass cpumasks by reference always.

Rather than passing cpumasks by value in all cases (which is
problematic for large NR_CPUS configurations), pass them 'by
reference' (i.e. through a pointer to a const cpumask).

On x86 this changes send_IPI_mask() to always only send IPIs to remote
CPUs (meaning any caller needing to handle the current CPU as well has
to do so on its own).

Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Keir Fraser <keir.fraser@citrix.com>

5 files changed, 34 insertions, 7 deletions

diff --git a/xen/common/Makefile b/xen/common/Makefile
index 08b9e2b00e..27b13f3e34 100644
--- a/xen/common/Makefile
+++ b/xen/common/Makefile
@@ -1,4 +1,5 @@
 obj-y += bitmap.o
+obj-y += cpu.o
 obj-y += domctl.o
 obj-y += domain.o
 obj-y += event_channel.o
diff --git a/xen/common/cpu.c b/xen/common/cpu.c
new file mode 100644
index 0000000000..96cba72e9e
--- /dev/null
+++ b/xen/common/cpu.c
@@ -0,0 +1,26 @@
+#include <xen/config.h>
+#include <xen/cpumask.h>
+
+/*
+ * cpu_bit_bitmap[] is a special, "compressed" data structure that
+ * represents all NR_CPUS bits binary values of 1<<nr.
+ *
+ * It is used by cpumask_of() to get a constant address to a CPU
+ * mask value that has a single bit set only.
+ */
+
+/* cpu_bit_bitmap[0] is empty - so we can back into it */
+#define MASK_DECLARE_1(x) [x+1][0] = 1UL << (x)
+#define MASK_DECLARE_2(x) MASK_DECLARE_1(x), MASK_DECLARE_1(x+1)
+#define MASK_DECLARE_4(x) MASK_DECLARE_2(x), MASK_DECLARE_2(x+2)
+#define MASK_DECLARE_8(x) MASK_DECLARE_4(x), MASK_DECLARE_4(x+4)
+
+const unsigned long cpu_bit_bitmap[BITS_PER_LONG+1][BITS_TO_LONGS(NR_CPUS)] = {
+
+    MASK_DECLARE_8(0),  MASK_DECLARE_8(8),
+    MASK_DECLARE_8(16), MASK_DECLARE_8(24),
+#if BITS_PER_LONG > 32
+    MASK_DECLARE_8(32), MASK_DECLARE_8(40),
+    MASK_DECLARE_8(48), MASK_DECLARE_8(56),
+#endif
+};
diff --git a/xen/common/grant_table.c b/xen/common/grant_table.c
index 530635469d..7bbc05d896 100644
--- a/xen/common/grant_table.c
+++ b/xen/common/grant_table.c
@@ -715,7 +715,7 @@ gnttab_unmap_grant_ref(
                 goto fault;
         }
 
-        flush_tlb_mask(current->domain->domain_dirty_cpumask);
+        flush_tlb_mask(&current->domain->domain_dirty_cpumask);
 
         for ( i = 0; i < partial_done; i++ )
             __gnttab_unmap_common_complete(&(common[i]));
@@ -727,7 +727,7 @@ gnttab_unmap_grant_ref(
     return 0;
 
 fault:
-    flush_tlb_mask(current->domain->domain_dirty_cpumask);
+    flush_tlb_mask(&current->domain->domain_dirty_cpumask);
 
     for ( i = 0; i < partial_done; i++ )
         __gnttab_unmap_common_complete(&(common[i]));
@@ -774,7 +774,7 @@ gnttab_unmap_and_replace(
                 goto fault;
         }
         
-        flush_tlb_mask(current->domain->domain_dirty_cpumask);
+        flush_tlb_mask(&current->domain->domain_dirty_cpumask);
         
         for ( i = 0; i < partial_done; i++ )
             __gnttab_unmap_common_complete(&(common[i]));
@@ -786,7 +786,7 @@ gnttab_unmap_and_replace(
     return 0;
 
 fault:
-    flush_tlb_mask(current->domain->domain_dirty_cpumask);
+    flush_tlb_mask(&current->domain->domain_dirty_cpumask);
 
     for ( i = 0; i < partial_done; i++ )
         __gnttab_unmap_common_complete(&(common[i]));
@@ -1123,7 +1123,7 @@ gnttab_transfer(
 #ifndef __ia64__ /* IA64 implicitly replaces the old page in steal_page(). */
         guest_physmap_remove_page(d, gop.mfn, mfn, 0);
 #endif
-        flush_tlb_mask(d->domain_dirty_cpumask);
+        flush_tlb_mask(&d->domain_dirty_cpumask);
 
         /* Find the target domain. */
         if ( unlikely((e = rcu_lock_domain_by_id(gop.domid)) == NULL) )
diff --git a/xen/common/keyhandler.c b/xen/common/keyhandler.c
index c481df0f3f..5407456290 100644
--- a/xen/common/keyhandler.c
+++ b/xen/common/keyhandler.c
@@ -119,7 +119,7 @@ static void dump_registers(unsigned char key, struct cpu_user_regs *regs)
         if ( cpu == smp_processor_id() )
             continue;
         printk("\n*** Dumping CPU%d host state: ***\n", cpu);
-        on_selected_cpus(cpumask_of_cpu(cpu), __dump_execstate, NULL, 1, 1);
+        on_selected_cpus(cpumask_of(cpu), __dump_execstate, NULL, 1, 1);
     }
 
     printk("\n");
diff --git a/xen/common/page_alloc.c b/xen/common/page_alloc.c
index bb143aedd6..bd514cfeed 100644
--- a/xen/common/page_alloc.c
+++ b/xen/common/page_alloc.c
@@ -431,7 +431,7 @@ static struct page_info *alloc_heap_pages(
     if ( unlikely(!cpus_empty(mask)) )
     {
         perfc_incr(need_flush_tlb_flush);
-        flush_tlb_mask(mask);
+        flush_tlb_mask(&mask);
     }
 
     return pg;