Hook improvements (take 2) (#563)

Hi @tiann.

Thanks for the great project, I had great fun playing around with it.

This PR mainly tries to further minimize the possible delays caused by
KernelSU hooking.

There are 3 major changes:
- Processes with 0 < UID < 2000 are blocked straight-up before going
through the allow_list.
I don't see any need for such processes to be interested in root, and
this allows returning early before going through a more expensive
lookup.
If there's an expected breakage due to this change, I'll remove it. Let
me know.
- A page-sized (4K) bitmap is added.
This allows O(1) lookup for UID <= 32767.
This speeds up `ksu_is_allow_uid()` by about 4.8x by sacrificing a 4K
memory. IMHO, a good trade-off.
Most notably, this reduces the 99.999% result previously from worrying
milliseconds scale to microseconds scale.
For UID > 32767, another page-sized (4K) sequential array is used to
cache allow_list.

Compared to the previous PR #557, this new approach gives another nice
25% performance boost in average, 63-96% boost in worst cases.

Benchmark results are available at
https://docs.google.com/spreadsheets/d/1w_tO1zRLPNMFRer49pL1TQfL6ndEhilRrDU1XFIcWXY/edit?usp=sharing

Thanks!

---------

Signed-off-by: Juhyung Park <qkrwngud825@gmail.com>

website/docs/guide/how-to-integrate-for-non-gki.md

@@ -75,14 +75,20 @@ index ac59664eaecf..bdd585e1d2cc 100644
  	return retval;
  }
  
++extern bool ksu_execveat_hook __read_mostly;
 +extern int ksu_handle_execveat(int *fd, struct filename **filename_ptr, void *argv,
 +			void *envp, int *flags);
++extern int ksu_handle_execveat_sucompat(int *fd, struct filename **filename_ptr,
++				 void *argv, void *envp, int *flags);
  static int do_execveat_common(int fd, struct filename *filename,
  			      struct user_arg_ptr argv,
  			      struct user_arg_ptr envp,
  			      int flags)
  {
-+	ksu_handle_execveat(&fd, &filename, &argv, &envp, &flags);
++	if (unlikely(ksu_execveat_hook))
++		ksu_handle_execveat(&fd, &filename, &argv, &envp, &flags);
++	else
++		ksu_handle_execveat_sucompat(&fd, &filename, &argv, &envp, &flags);
  	return __do_execve_file(fd, filename, argv, envp, flags, NULL);
  }
  
@@ -115,14 +121,16 @@ index 650fc7e0f3a6..55be193913b6 100644
  }
  EXPORT_SYMBOL(kernel_read);
  
++extern bool ksu_vfs_read_hook __read_mostly;
 +extern int ksu_handle_vfs_read(struct file **file_ptr, char __user **buf_ptr,
 +			size_t *count_ptr, loff_t **pos);
  ssize_t vfs_read(struct file *file, char __user *buf, size_t count, loff_t *pos)
  {
  	ssize_t ret;
  
-+	ksu_handle_vfs_read(&file, &buf, &count, &pos);
-+	
++	if (unlikely(ksu_vfs_read_hook))
++		ksu_handle_vfs_read(&file, &buf, &count, &pos);
++
  	if (!(file->f_mode & FMODE_READ))
  		return -EBADF;
  	if (!(file->f_mode & FMODE_CAN_READ))
@@ -222,19 +230,22 @@ index 45306f9ef247..815091ebfca4 100755
 --- a/drivers/input/input.c
 +++ b/drivers/input/input.c
 @@ -367,10 +367,13 @@ static int input_get_disposition(struct input_dev *dev,
-        return disposition;
+ 	return disposition;
  }
-
+ 
++extern bool ksu_input_hook __read_mostly;
 +extern int ksu_handle_input_handle_event(unsigned int *type, unsigned int *code, int *value);
 +
  static void input_handle_event(struct input_dev *dev,
-                               unsigned int type, unsigned int code, int value)
+ 			       unsigned int type, unsigned int code, int value)
  {
-        int disposition = input_get_disposition(dev, type, code, &value);
-+       ksu_handle_input_handle_event(&type, &code, &value);
-
-        if (disposition != INPUT_IGNORE_EVENT && type != EV_SYN)
-                add_input_randomness(type, code, value);
+	int disposition = input_get_disposition(dev, type, code, &value);
++
++	if (unlikely(ksu_input_hook))
++		ksu_handle_input_handle_event(&type, &code, &value);
+ 
+ 	if (disposition != INPUT_IGNORE_EVENT && type != EV_SYN)
+ 		add_input_randomness(type, code, value);
 ```
 
 Finally, build your kernel again, KernelSU should works well.