x86, bts: memory accounting

Impact: move the BTS buffer accounting to the mlock bucket

Add alloc_locked_buffer() and free_locked_buffer() functions to mm/mlock.c
to kalloc a buffer and account the locked memory to current.

Account the memory for the BTS buffer to the tracer.

Signed-off-by: Markus Metzger <markus.t.metzger@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

3 files changed, 81 insertions, 11 deletions

diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c
index 6ad2bb607650..0a5df5f82fb9 100644
--- a/arch/x86/kernel/ptrace.c
+++ b/arch/x86/kernel/ptrace.c
@@ -650,6 +650,24 @@ static int ptrace_bts_drain(struct task_struct *child,
         return drained;
 }
 
+static int ptrace_bts_allocate_buffer(struct task_struct *child, size_t size)
+{
+        child->bts_buffer = alloc_locked_buffer(size);
+        if (!child->bts_buffer)
+                return -ENOMEM;
+
+        child->bts_size = size;
+
+        return 0;
+}
+
+static void ptrace_bts_free_buffer(struct task_struct *child)
+{
+        free_locked_buffer(child->bts_buffer, child->bts_size);
+        child->bts_buffer = NULL;
+        child->bts_size = 0;
+}
+
 static int ptrace_bts_config(struct task_struct *child,
                              long cfg_size,
                              const struct ptrace_bts_config __user *ucfg)
@@ -679,14 +697,13 @@ static int ptrace_bts_config(struct task_struct *child,
 
         if ((cfg.flags & PTRACE_BTS_O_ALLOC) &&
             (cfg.size != child->bts_size)) {
-                kfree(child->bts_buffer);
+                int error;
 
-                child->bts_size = cfg.size;
-                child->bts_buffer = kzalloc(cfg.size, GFP_KERNEL);
-                if (!child->bts_buffer) {
-                        child->bts_size = 0;
-                        return -ENOMEM;
-                }
+                ptrace_bts_free_buffer(child);
+
+                error = ptrace_bts_allocate_buffer(child, cfg.size);
+                if (error < 0)
+                        return error;
         }
 
         if (cfg.flags & PTRACE_BTS_O_TRACE)
@@ -701,10 +718,8 @@ static int ptrace_bts_config(struct task_struct *child,
         if (IS_ERR(child->bts)) {
                 int error = PTR_ERR(child->bts);
 
-                kfree(child->bts_buffer);
+                ptrace_bts_free_buffer(child);
                 child->bts = NULL;
-                child->bts_buffer = NULL;
-                child->bts_size = 0;
 
                 return error;
         }
@@ -784,6 +799,9 @@ static void ptrace_bts_untrace(struct task_struct *child)
                 ds_release_bts(child->bts);
                 child->bts = NULL;
 
+                /* We cannot update total_vm and locked_vm since
+                   child's mm is already gone. But we can reclaim the
+                   memory. */
                 kfree(child->bts_buffer);
                 child->bts_buffer = NULL;
                 child->bts_size = 0;
@@ -792,7 +810,12 @@ static void ptrace_bts_untrace(struct task_struct *child)
 
 static void ptrace_bts_detach(struct task_struct *child)
 {
-        ptrace_bts_untrace(child);
+        if (unlikely(child->bts)) {
+                ds_release_bts(child->bts);
+                child->bts = NULL;
+
+                ptrace_bts_free_buffer(child);
+        }
 }
 #else
 static inline void ptrace_bts_fork(struct task_struct *tsk) {}
diff --git a/include/linux/mm.h b/include/linux/mm.h
index ffee2f743418..9979d3fab6e7 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1286,5 +1286,7 @@ int vmemmap_populate_basepages(struct page *start_page,
 int vmemmap_populate(struct page *start_page, unsigned long pages, int node);
 void vmemmap_populate_print_last(void);
 
+extern void *alloc_locked_buffer(size_t size);
+extern void free_locked_buffer(void *buffer, size_t size);
 #endif /* __KERNEL__ */
 #endif /* _LINUX_MM_H */
diff --git a/mm/mlock.c b/mm/mlock.c
index 1ada366570cb..3035a56e7616 100644
--- a/mm/mlock.c
+++ b/mm/mlock.c
@@ -667,3 +667,48 @@ void user_shm_unlock(size_t size, struct user_struct *user)
         spin_unlock(&shmlock_user_lock);
         free_uid(user);
 }
+
+void *alloc_locked_buffer(size_t size)
+{
+        unsigned long rlim, vm, pgsz;
+        void *buffer = NULL;
+
+        pgsz = PAGE_ALIGN(size) >> PAGE_SHIFT;
+
+        down_write(&current->mm->mmap_sem);
+
+        rlim = current->signal->rlim[RLIMIT_AS].rlim_cur >> PAGE_SHIFT;
+        vm   = current->mm->total_vm + pgsz;
+        if (rlim < vm)
+                goto out;
+
+        rlim = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur >> PAGE_SHIFT;
+        vm   = current->mm->locked_vm + pgsz;
+        if (rlim < vm)
+                goto out;
+
+        buffer = kzalloc(size, GFP_KERNEL);
+        if (!buffer)
+                goto out;
+
+        current->mm->total_vm  += pgsz;
+        current->mm->locked_vm += pgsz;
+
+ out:
+        up_write(&current->mm->mmap_sem);
+        return buffer;
+}
+
+void free_locked_buffer(void *buffer, size_t size)
+{
+        unsigned long pgsz = PAGE_ALIGN(size) >> PAGE_SHIFT;
+
+        down_write(&current->mm->mmap_sem);
+
+        current->mm->total_vm  -= pgsz;
+        current->mm->locked_vm -= pgsz;
+
+        up_write(&current->mm->mmap_sem);
+
+        kfree(buffer);
+}