All of lore.kernel.org
 help / color / mirror / Atom feed
* [sparc32] vm fault code handling
@ 2004-12-22 14:31 William Lee Irwin III
  2004-12-22 23:38 ` Art Haas
  2004-12-23  3:23 ` William Lee Irwin III
  0 siblings, 2 replies; 3+ messages in thread
From: William Lee Irwin III @ 2004-12-22 14:31 UTC (permalink / raw)
  To: sparclinux

Convert arch/sparc/mm/fault.c to symbolic fault codes, default to minor
faults for unrecognized fault codes. Also treat VM_FAULT_OOM similarly
to VM_FAULT_SIGBUS in force_user_fault(), thereby fixing a latent bug.

Index: sparc32-2.6.10-rc3/arch/sparc/mm/fault.c
=================================--- sparc32-2.6.10-rc3.orig/arch/sparc/mm/fault.c	2004-10-07 04:01:27.000000000 -0700
+++ sparc32-2.6.10-rc3/arch/sparc/mm/fault.c	2004-12-21 08:10:56.962777738 -0800
@@ -294,16 +294,17 @@
 	 * the fault.
 	 */
 	switch (handle_mm_fault(mm, vma, address, write)) {
-	case 1:
-		current->min_flt++;
-		break;
-	case 2:
+	case VM_FAULT_SIGBUS:
+		goto do_sigbus;
+	case VM_FAULT_OOM:
+		goto out_of_memory;
+	case VM_FAULT_MAJOR:
 		current->maj_flt++;
 		break;
-	case 0:
-		goto do_sigbus;
+	case VM_FAULT_MINOR:
 	default:
-		goto out_of_memory;
+		current->min_flt++;
+		break;
 	}
 	up_read(&mm->mmap_sem);
 	return;
@@ -535,8 +536,11 @@
 		if(!(vma->vm_flags & (VM_READ | VM_EXEC)))
 			goto bad_area;
 	}
-	if (!handle_mm_fault(mm, vma, address, write))
+	switch (handle_mm_fault(mm, vma, address, write)) {
+	case VM_FAULT_SIGBUS:
+	case VM_FAULT_OOM:
 		goto do_sigbus;
+	}
 	up_read(&mm->mmap_sem);
 	return;
 bad_area:

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [sparc32] vm fault code handling
  2004-12-22 14:31 [sparc32] vm fault code handling William Lee Irwin III
@ 2004-12-22 23:38 ` Art Haas
  2004-12-23  3:23 ` William Lee Irwin III
  1 sibling, 0 replies; 3+ messages in thread
From: Art Haas @ 2004-12-22 23:38 UTC (permalink / raw)
  To: sparclinux

On Wed, Dec 22, 2004 at 06:31:13AM -0800, William Lee Irwin III wrote:
> Convert arch/sparc/mm/fault.c to symbolic fault codes, default to minor
> faults for unrecognized fault codes. Also treat VM_FAULT_OOM similarly
> to VM_FAULT_SIGBUS in force_user_fault(), thereby fixing a latent bug.
> 
> [ ... snip patch ... ]

Hi.

Here's a slightly modified version of that patch which makes the sparc32
code more closely match the sparc64 code by setting the default branch
of the case statement to BUG(). The rest of the patch matched what was
posted above.

Art Haas

=== arch/sparc/mm/fault.c 1.19 vs edited ==--- 1.19/arch/sparc/mm/fault.c	2004-07-13 07:56:23 -05:00
+++ edited/arch/sparc/mm/fault.c	2004-12-22 17:24:37 -06:00
@@ -294,16 +294,18 @@
 	 * the fault.
 	 */
 	switch (handle_mm_fault(mm, vma, address, write)) {
-	case 1:
-		current->min_flt++;
-		break;
-	case 2:
+	case VM_FAULT_SIGBUS:
+		goto do_sigbus;
+	case VM_FAULT_OOM:
+		goto out_of_memory;
+	case VM_FAULT_MAJOR:
 		current->maj_flt++;
 		break;
-	case 0:
-		goto do_sigbus;
+	case VM_FAULT_MINOR:
+		current->min_flt++;
+		break;
 	default:
-		goto out_of_memory;
+		BUG();
 	}
 	up_read(&mm->mmap_sem);
 	return;
@@ -535,8 +537,11 @@
 		if(!(vma->vm_flags & (VM_READ | VM_EXEC)))
 			goto bad_area;
 	}
-	if (!handle_mm_fault(mm, vma, address, write))
+	switch (handle_mm_fault(mm, vma, address, write)) {
+	case VM_FAULT_SIGBUS:
+	case VM_FAULT_OOM:
 		goto do_sigbus;
+	}
 	up_read(&mm->mmap_sem);
 	return;
 bad_area:


-- 
Man once surrendering his reason, has no remaining guard against absurdities
the most monstrous, and like a ship without rudder, is the sport of every wind.

-Thomas Jefferson to James Smith, 1822

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [sparc32] vm fault code handling
  2004-12-22 14:31 [sparc32] vm fault code handling William Lee Irwin III
  2004-12-22 23:38 ` Art Haas
@ 2004-12-23  3:23 ` William Lee Irwin III
  1 sibling, 0 replies; 3+ messages in thread
From: William Lee Irwin III @ 2004-12-23  3:23 UTC (permalink / raw)
  To: sparclinux

On Wed, Dec 22, 2004 at 06:31:13AM -0800, William Lee Irwin III wrote:
>> Convert arch/sparc/mm/fault.c to symbolic fault codes, default to minor
>> faults for unrecognized fault codes. Also treat VM_FAULT_OOM similarly
>> to VM_FAULT_SIGBUS in force_user_fault(), thereby fixing a latent bug.
>> 
>> [ ... snip patch ... ]

On Wed, Dec 22, 2004 at 05:38:36PM -0600, Art Haas wrote:
> Here's a slightly modified version of that patch which makes the sparc32
> code more closely match the sparc64 code by setting the default branch
> of the case statement to BUG(). The rest of the patch matched what was
> posted above.

I considered that, flipped a coin, and went with defaulting to minor
faults. Since someone has an opinion (though not necessarily strong)
I'll switch it over to that way.


-- wli

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2004-12-23  3:23 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-12-22 14:31 [sparc32] vm fault code handling William Lee Irwin III
2004-12-22 23:38 ` Art Haas
2004-12-23  3:23 ` William Lee Irwin III

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.