[PATCH] 2.5.20 x86 iobitmap cleanup

This makes the IO bitmap allocations a separately allocated structure,
shrinking the default task size.

We alloc it in sys_ioperm() and copy_thread() and frees in
exit_thread().  It also gets rid of the IO_BITMAP_SIZE+1 crap, as only
the tss actually needs the tail long, and we weren't copying it into the
bitmap anyways.

arch/i386/kernel/ioport.c

@@ -10,10 +10,10 @@
 #include <linux/errno.h>
 #include <linux/types.h>
 #include <linux/ioport.h>
-#include <linux/mm.h>
 #include <linux/smp.h>
 #include <linux/smp_lock.h>
 #include <linux/stddef.h>
+#include <linux/slab.h>
 
 /* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
 static void set_bitmap(unsigned long *bitmap, short base, short extent, int new_value)
@@ -66,12 +66,16 @@ asmlinkage int sys_ioperm(unsigned long from, unsigned long num, int turn_on)
 	 * IO bitmap up. ioperm() is much less timing critical than clone(),
 	 * this is why we delay this operation until now:
 	 */
-	if (!t->ioperm) {
+	if (!t->ts_io_bitmap) {
+		unsigned long *bitmap;
+		bitmap = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
+		if (!bitmap)
+			return -ENOMEM;
 		/*
 		 * just in case ...
 		 */
-		memset(t->io_bitmap,0xff,(IO_BITMAP_SIZE+1)*4);
-		t->ioperm = 1;
+		memset(bitmap, 0xff, IO_BITMAP_BYTES);
+		t->ts_io_bitmap = bitmap;
 		/*
 		 * this activates it in the TSS
 		 */
@@ -81,7 +85,7 @@ asmlinkage int sys_ioperm(unsigned long from, unsigned long num, int turn_on)
 	/*
 	 * do it in the per-thread copy and in the TSS ...
 	 */
-	set_bitmap(t->io_bitmap, from, num, !turn_on);
+	set_bitmap(t->ts_io_bitmap, from, num, !turn_on);
 	set_bitmap(tss->io_bitmap, from, num, !turn_on);
 
 	return 0;

arch/i386/kernel/process.c

@@ -509,7 +509,13 @@ int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
  */
 void exit_thread(void)
 {
-	/* nothing to do ... */
+	struct task_struct *tsk = current;
+
+	/* The process may have allocated an io port bitmap... nuke it. */
+	if (unlikely(NULL != tsk->thread.ts_io_bitmap)) {
+		kfree(tsk->thread.ts_io_bitmap);
+		tsk->thread.ts_io_bitmap = NULL;
+	}
 }
 
 void flush_thread(void)
@@ -549,6 +555,7 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long esp,
 	struct task_struct * p, struct pt_regs * regs)
 {
 	struct pt_regs * childregs;
+	struct task_struct *tsk;
 
 	childregs = ((struct pt_regs *) (THREAD_SIZE + (unsigned long) p->thread_info)) - 1;
 	struct_cpy(childregs, regs);
@@ -563,8 +570,17 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long esp,
 	savesegment(fs,p->thread.fs);
 	savesegment(gs,p->thread.gs);
 
-	unlazy_fpu(current);
-	struct_cpy(&p->thread.i387, &current->thread.i387);
+	tsk = current;
+	unlazy_fpu(tsk);
+	struct_cpy(&p->thread.i387, &tsk->thread.i387);
+
+	if (unlikely(NULL != tsk->thread.ts_io_bitmap)) {
+		p->thread.ts_io_bitmap = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL);
+		if (!p->thread.ts_io_bitmap)
+			return -ENOMEM;
+		memcpy(p->thread.ts_io_bitmap, tsk->thread.ts_io_bitmap,
+			IO_BITMAP_BYTES);
+	}
 
 	return 0;
 }
@@ -685,8 +701,8 @@ void __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
 		loaddebug(next, 7);
 	}
 
-	if (unlikely(prev->ioperm || next->ioperm)) {
-		if (next->ioperm) {
+	if (unlikely(prev->ts_io_bitmap || next->ts_io_bitmap)) {
+		if (next->ts_io_bitmap) {
 			/*
 			 * 4 cachelines copy ... not good, but not that
 			 * bad either. Anyone got something better?
@@ -695,8 +711,8 @@ void __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
 			 * and playing VM tricks to switch the IO bitmap
 			 * is not really acceptable.]
 			 */
-			memcpy(tss->io_bitmap, next->io_bitmap,
-				 IO_BITMAP_SIZE*sizeof(unsigned long));
+			memcpy(tss->io_bitmap, next->ts_io_bitmap,
+				IO_BITMAP_BYTES);
 			tss->bitmap = IO_BITMAP_OFFSET;
 		} else
 			/*

include/asm-i386/processor.h

@@ -267,6 +267,7 @@ extern unsigned int mca_pentium_flag;
  * Size of io_bitmap in longwords: 32 is ports 0-0x3ff.
  */
 #define IO_BITMAP_SIZE	32
+#define IO_BITMAP_BYTES	(IO_BITMAP_SIZE * 4)
 #define IO_BITMAP_OFFSET offsetof(struct tss_struct,io_bitmap)
 #define INVALID_IO_BITMAP_OFFSET 0x8000
 
@@ -370,8 +371,7 @@ struct thread_struct {
 	unsigned long		screen_bitmap;
 	unsigned long		v86flags, v86mask, v86mode, saved_esp0;
 /* IO permissions */
-	int		ioperm;
-	unsigned long	io_bitmap[IO_BITMAP_SIZE+1];
+	unsigned long	*ts_io_bitmap;
 };
 
 #define INIT_THREAD  {						\
@@ -381,7 +381,7 @@ struct thread_struct {
 	0, 0, 0,						\
 	{ { 0, }, },		/* 387 state */			\
 	0,0,0,0,0,0,						\
-	0,{~0,}			/* io permissions */		\
+	NULL,			/* io permissions */		\
 }
 
 #define INIT_TSS  {						\