[PATCH] MTRR suspend/resume cleanup

There has been some discuss about solving the SMP MTRR suspend/resume
breakage, but I didn't find a patch for it.  This is an intent for it.  The
basic idea is moving mtrr initializing into cpu_identify for all APs (so it
works for cpu hotplug).  For BP, restore_processor_state is responsible for
restoring MTRR.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

5 files changed, 70 insertions, 35 deletions

diff --git a/arch/i386/kernel/cpu/common.c b/arch/i386/kernel/cpu/common.c
index 2203a9d20212..4553ffd94b1f 100644
--- a/arch/i386/kernel/cpu/common.c
+++ b/arch/i386/kernel/cpu/common.c
@@ -435,6 +435,11 @@ void __devinit identify_cpu(struct cpuinfo_x86 *c)
         if (c == &boot_cpu_data)
                 sysenter_setup();
         enable_sep_cpu();
+
+        if (c == &boot_cpu_data)
+                mtrr_bp_init();
+        else
+                mtrr_ap_init();
 }
 
 #ifdef CONFIG_X86_HT
diff --git a/arch/i386/kernel/cpu/mtrr/generic.c b/arch/i386/kernel/cpu/mtrr/generic.c
index 64d91f73a0a4..169ac8e0db68 100644
--- a/arch/i386/kernel/cpu/mtrr/generic.c
+++ b/arch/i386/kernel/cpu/mtrr/generic.c
@@ -67,13 +67,6 @@ void __init get_mtrr_state(void)
         mtrr_state.enabled = (lo & 0xc00) >> 10;
 }
 
-/*  Free resources associated with a struct mtrr_state  */
-void __init finalize_mtrr_state(void)
-{
-        kfree(mtrr_state.var_ranges);
-        mtrr_state.var_ranges = NULL;
-}
-
 /*  Some BIOS's are fucked and don't set all MTRRs the same!  */
 void __init mtrr_state_warn(void)
 {
@@ -334,6 +327,9 @@ static void generic_set_mtrr(unsigned int reg, unsigned long base,
 */
 {
         unsigned long flags;
+        struct mtrr_var_range *vr;
+
+        vr = &mtrr_state.var_ranges[reg];
 
         local_irq_save(flags);
         prepare_set();
@@ -342,11 +338,15 @@ static void generic_set_mtrr(unsigned int reg, unsigned long base,
                 /* The invalid bit is kept in the mask, so we simply clear the
                    relevant mask register to disable a range. */
                 mtrr_wrmsr(MTRRphysMask_MSR(reg), 0, 0);
+                memset(vr, 0, sizeof(struct mtrr_var_range));
         } else {
-                mtrr_wrmsr(MTRRphysBase_MSR(reg), base << PAGE_SHIFT | type,
-                      (base & size_and_mask) >> (32 - PAGE_SHIFT));
-                mtrr_wrmsr(MTRRphysMask_MSR(reg), -size << PAGE_SHIFT | 0x800,
-                      (-size & size_and_mask) >> (32 - PAGE_SHIFT));
+                vr->base_lo = base << PAGE_SHIFT | type;
+                vr->base_hi = (base & size_and_mask) >> (32 - PAGE_SHIFT);
+                vr->mask_lo = -size << PAGE_SHIFT | 0x800;
+                vr->mask_hi = (-size & size_and_mask) >> (32 - PAGE_SHIFT);
+
+                mtrr_wrmsr(MTRRphysBase_MSR(reg), vr->base_lo, vr->base_hi);
+                mtrr_wrmsr(MTRRphysMask_MSR(reg), vr->mask_lo, vr->mask_hi);
         }
 
         post_set();
diff --git a/arch/i386/kernel/cpu/mtrr/main.c b/arch/i386/kernel/cpu/mtrr/main.c
index d66b09e0c820..764cac64e211 100644
--- a/arch/i386/kernel/cpu/mtrr/main.c
+++ b/arch/i386/kernel/cpu/mtrr/main.c
@@ -332,6 +332,8 @@ int mtrr_add_page(unsigned long base, unsigned long size,
 
         error = -EINVAL;
 
+        /* No CPU hotplug when we change MTRR entries */
+        lock_cpu_hotplug();
         /*  Search for existing MTRR  */
         down(&main_lock);
         for (i = 0; i < num_var_ranges; ++i) {
@@ -372,6 +374,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
         error = i;
  out:
         up(&main_lock);
+        unlock_cpu_hotplug();
         return error;
 }
 
@@ -461,6 +464,8 @@ int mtrr_del_page(int reg, unsigned long base, unsigned long size)
                 return -ENXIO;
 
         max = num_var_ranges;
+        /* No CPU hotplug when we change MTRR entries */
+        lock_cpu_hotplug();
         down(&main_lock);
         if (reg < 0) {
                 /*  Search for existing MTRR  */
@@ -501,6 +506,7 @@ int mtrr_del_page(int reg, unsigned long base, unsigned long size)
         error = reg;
  out:
         up(&main_lock);
+        unlock_cpu_hotplug();
         return error;
 }
 /**
@@ -544,21 +550,9 @@ static void __init init_ifs(void)
         centaur_init_mtrr();
 }
 
-static void __init init_other_cpus(void)
-{
-        if (use_intel())
-                get_mtrr_state();
-
-        /* bring up the other processors */
-        set_mtrr(~0U,0,0,0);
-
-        if (use_intel()) {
-                finalize_mtrr_state();
-                mtrr_state_warn();
-        }
-}
-
-
+/* The suspend/resume methods are only for CPU without MTRR. CPU using generic
+ * MTRR driver doesn't require this
+ */
 struct mtrr_value {
         mtrr_type       ltype;
         unsigned long   lbase;
@@ -611,13 +605,13 @@ static struct sysdev_driver mtrr_sysdev_driver = {
 
 
 /**
- * mtrr_init - initialize mtrrs on the boot CPU
+ * mtrr_bp_init - initialize mtrrs on the boot CPU
  *
  * This needs to be called early; before any of the other CPUs are 
  * initialized (i.e. before smp_init()).
  * 
  */
-static int __init mtrr_init(void)
+void __init mtrr_bp_init(void)
 {
         init_ifs();
 
@@ -674,12 +668,48 @@ static int __init mtrr_init(void)
         if (mtrr_if) {
                 set_num_var_ranges();
                 init_table();
-                init_other_cpus();
-
-                return sysdev_driver_register(&cpu_sysdev_class,
-                                              &mtrr_sysdev_driver);
+                if (use_intel())
+                        get_mtrr_state();
         }
-        return -ENXIO;
 }
 
-subsys_initcall(mtrr_init);
+void mtrr_ap_init(void)
+{
+        unsigned long flags;
+
+        if (!mtrr_if || !use_intel())
+                return;
+        /*
+         * Ideally we should hold main_lock here to avoid mtrr entries changed,
+         * but this routine will be called in cpu boot time, holding the lock
+         * breaks it. This routine is called in two cases: 1.very earily time
+         * of software resume, when there absolutely isn't mtrr entry changes;
+         * 2.cpu hotadd time. We let mtrr_add/del_page hold cpuhotplug lock to
+         * prevent mtrr entry changes
+         */
+        local_irq_save(flags);
+
+        mtrr_if->set_all();
+
+        local_irq_restore(flags);
+}
+
+static int __init mtrr_init_finialize(void)
+{
+        if (!mtrr_if)
+                return 0;
+        if (use_intel())
+                mtrr_state_warn();
+        else {
+                /* The CPUs haven't MTRR and seemes not support SMP. They have
+                 * specific drivers, we use a tricky method to support
+                 * suspend/resume for them.
+                 * TBD: is there any system with such CPU which supports
+                 * suspend/resume?  if no, we should remove the code.
+                 */
+                sysdev_driver_register(&cpu_sysdev_class,
+                        &mtrr_sysdev_driver);
+        }
+        return 0;
+}
+subsys_initcall(mtrr_init_finialize);
diff --git a/arch/i386/kernel/cpu/mtrr/mtrr.h b/arch/i386/kernel/cpu/mtrr/mtrr.h
index de1351245599..99c9f2682041 100644
--- a/arch/i386/kernel/cpu/mtrr/mtrr.h
+++ b/arch/i386/kernel/cpu/mtrr/mtrr.h
@@ -91,7 +91,6 @@ extern struct mtrr_ops * mtrr_if;
 
 extern unsigned int num_var_ranges;
 
-void finalize_mtrr_state(void);
 void mtrr_state_warn(void);
 char *mtrr_attrib_to_str(int x);
 void mtrr_wrmsr(unsigned, unsigned, unsigned);
diff --git a/arch/i386/power/cpu.c b/arch/i386/power/cpu.c
index 0e6b45b61251..c547c1af6fa1 100644
--- a/arch/i386/power/cpu.c
+++ b/arch/i386/power/cpu.c
@@ -137,6 +137,7 @@ void __restore_processor_state(struct saved_context *ctxt)
 
         fix_processor_context();
         do_fpu_end();
+        mtrr_ap_init();
 }
 
 void restore_processor_state(void)