x86: possibly bring up all CPUs even if not all are supposed to be used

Reportedly Intel CPUs which can't broadcast #MC to all targeted
cores/threads because some have CR4.MCE clear will shut down. Therefore
we want to keep CR4.MCE enabled when offlining a CPU, and we need to
bring up all CPUs in order to be able to set CR4.MCE in the first place.

The use of clear_in_cr4() in cpu_mcheck_disable() was ill advised
anyway, and to avoid future similar mistakes I'm removing clear_in_cr4()
altogether right here.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Wei Liu <wei.liu2@citrix.com>
master commit: 8797d20a6ec2dd75195585a107ce345c51c0a59a
master date: 2018-07-19 13:43:33 +0100

diff --git a/xen/arch/x86/cpu/common.c b/xen/arch/x86/cpu/common.c
index 147c83d94242c88d2e1fa91474f415fe1e5c6559..fc9e379096c8291963bdd95de4ac7cd57736cad7 100644 (file)
--- a/xen/arch/x86/cpu/common.c
+++ b/xen/arch/x86/cpu/common.c
@@ -15,6 +15,7 @@
 #include <public/sysctl.h> /* for XEN_INVALID_{SOCKET,CORE}_ID */
 
 #include "cpu.h"
+#include "mcheck/x86_mca.h"
 
 bool_t opt_arat = 1;
 boolean_param("arat", opt_arat);
@@ -252,6 +253,9 @@ static void __init early_cpu_detect(void)
                        hap_paddr_bits = PADDR_BITS;
        }
 
+       if (c->x86_vendor != X86_VENDOR_AMD)
+               park_offline_cpus = opt_mce;
+
        initialize_cpu_data(0);
 }
 

diff --git a/xen/arch/x86/cpu/mcheck/mce_intel.c b/xen/arch/x86/cpu/mcheck/mce_intel.c
index 005e41de60d1f09e749c48e0b9087f8ff2331ff3..71e6891be1ea9a9ac5dae8c20863156cd132d4e4 100644 (file)
--- a/xen/arch/x86/cpu/mcheck/mce_intel.c
+++ b/xen/arch/x86/cpu/mcheck/mce_intel.c
@@ -628,8 +628,6 @@ static void clear_cmci(void)
 
 static void cpu_mcheck_disable(void)
 {
-    clear_in_cr4(X86_CR4_MCE);
-
     if (cmci_support && opt_mce)
         clear_cmci();
 }

diff --git a/xen/arch/x86/mpparse.c b/xen/arch/x86/mpparse.c
index ef6557c0ef7ae9114134efbfd2dc1596550702a7..31e29d90c2d51f54c2fdd240ed4eebb1ec44f9df 100644 (file)
--- a/xen/arch/x86/mpparse.c
+++ b/xen/arch/x86/mpparse.c
@@ -69,19 +69,26 @@ physid_mask_t phys_cpu_present_map;
 
 void __init set_nr_cpu_ids(unsigned int max_cpus)
 {
+       unsigned int tot_cpus = num_processors + disabled_cpus;
+
        if (!max_cpus)
-               max_cpus = num_processors + disabled_cpus;
+               max_cpus = tot_cpus;
        if (max_cpus > NR_CPUS)
                max_cpus = NR_CPUS;
        else if (!max_cpus)
                max_cpus = 1;
        printk(XENLOG_INFO "SMP: Allowing %u CPUs (%d hotplug CPUs)\n",
               max_cpus, max_t(int, max_cpus - num_processors, 0));
-       nr_cpu_ids = max_cpus;
+
+       if (!park_offline_cpus)
+               tot_cpus = max_cpus;
+       nr_cpu_ids = min(tot_cpus, NR_CPUS + 0u);
+       if (park_offline_cpus && nr_cpu_ids < num_processors)
+               printk(XENLOG_WARNING "SMP: Cannot bring up %u further CPUs\n",
+                      num_processors - nr_cpu_ids);
 
 #ifndef nr_cpumask_bits
-       nr_cpumask_bits = (max_cpus + (BITS_PER_LONG - 1)) &
-                         ~(BITS_PER_LONG - 1);
+       nr_cpumask_bits = ROUNDUP(nr_cpu_ids, BITS_PER_LONG);
        printk(XENLOG_DEBUG "NR_CPUS:%u nr_cpumask_bits:%u\n",
               NR_CPUS, nr_cpumask_bits);
 #endif

diff --git a/xen/arch/x86/setup.c b/xen/arch/x86/setup.c
index ceb31c2cd127a8694ab0ea0dffee915600d8275e..f3f751a8937bdcf0754a32b56b1848b118d7bc8e 100644 (file)
--- a/xen/arch/x86/setup.c
+++ b/xen/arch/x86/setup.c
@@ -602,7 +602,7 @@ void __init noreturn __start_xen(unsigned long mbi_p)
 {
     char *memmap_type = NULL;
     char *cmdline, *kextra, *loader;
-    unsigned int initrdidx, domcr_flags = DOMCRF_s3_integrity;
+    unsigned int initrdidx, num_parked = 0, domcr_flags = DOMCRF_s3_integrity;
     multiboot_info_t *mbi = __va(mbi_p);
     module_t *mod = (module_t *)__va(mbi->mods_addr);
     unsigned long nr_pages, raw_max_page, modules_headroom, *module_map;
@@ -1379,7 +1379,8 @@ void __init noreturn __start_xen(unsigned long mbi_p)
     else
     {
         set_nr_cpu_ids(max_cpus);
-        max_cpus = nr_cpu_ids;
+        if ( !max_cpus )
+            max_cpus = nr_cpu_ids;
     }
 
     /* Low mappings were only needed for some BIOS table parsing. */
@@ -1488,15 +1489,26 @@ void __init noreturn __start_xen(unsigned long mbi_p)
         /* Set up node_to_cpumask based on cpu_to_node[]. */
         numa_add_cpu(i);        
 
-        if ( (num_online_cpus() < max_cpus) && !cpu_online(i) )
+        if ( (park_offline_cpus || num_online_cpus() < max_cpus) &&
+             !cpu_online(i) )
         {
             int ret = cpu_up(i);
             if ( ret != 0 )
                 printk("Failed to bring up CPU %u (error %d)\n", i, ret);
+            else if ( num_online_cpus() > max_cpus )
+            {
+                ret = cpu_down(i);
+                if ( !ret )
+                    ++num_parked;
+                else
+                    printk("Could not re-offline CPU%u (%d)\n", i, ret);
+            }
         }
     }
 
     printk("Brought up %ld CPUs\n", (long)num_online_cpus());
+    if ( num_parked )
+        printk(XENLOG_INFO "Parked %u CPUs\n", num_parked);
     smp_cpus_done();
 
     do_initcalls();

diff --git a/xen/include/asm-x86/processor.h b/xen/include/asm-x86/processor.h
index 1f4a5499fcd41b10765410c61fe3562d98a0b865..14bed92b9b95ea985bbe1d2086e142712d17e7f0 100644 (file)
--- a/xen/include/asm-x86/processor.h
+++ b/xen/include/asm-x86/processor.h
@@ -395,12 +395,6 @@ static always_inline void set_in_cr4 (unsigned long mask)
     write_cr4(read_cr4() | mask);
 }
 
-static always_inline void clear_in_cr4 (unsigned long mask)
-{
-    mmu_cr4_features &= ~mask;
-    write_cr4(read_cr4() & ~mask);
-}
-
 static inline unsigned int read_pkru(void)
 {
     unsigned int pkru;