Skip to contentSkip to conversation

ide-disk oopses on boot

PATCH In review Petr Vandrovec 4 participants 6 messages Aug 9, 2005 – Aug 10, 2005 1 message not in the archive

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ide-disk oopses on boot

Hello Christoph,
back in June your '[PATCH] NUMA aware block device control structure
allocation' patch went in, changing ide-disk.c's code:

- g = alloc_disk(1 << PARTN_BITS);
+ g = alloc_disk_node(1 << PARTN_BITS,
+ pcibus_to_node(drive->hwif->pci_dev->bus));

Problem is that pci_dev may be NULL - and it is NULL for example with
kernel I've just built, with amd IDE driver built as a module while with ide-disk built into the kernel.

I think that you probably want to guard your code by
'if (drive->hwif->pci_dev)', as besides my silly configuration mistake also ISA devices have a chance to have pci_dev NULL. Not that there are any such users if nobody hit it in the last 6 weeks...

I've just built amd IDE driver into the kernel, where it belongs anyway,
but just in case please apply this patch...
Thanks,
Petr Vandrovec

drivers/ide/ide-disk.c +6 -2
@@ -1219,8 +1219,12 @@
 	if (!idkp)
 		goto failed;
 
-	g = alloc_disk_node(1 << PARTN_BITS,
-			pcibus_to_node(drive->hwif->pci_dev->bus));
+	if (drive->hwif->pci_dev) {
+		g = alloc_disk_node(1 << PARTN_BITS,
+				pcibus_to_node(drive->hwif->pci_dev->bus));
+	} else {
+		g = alloc_disk(1 << PARTN_BITS);
+	}
 	if (!g)
 		goto out_free_idkp;
 
#1 lore raw
Quoted 3 lines · Petr Vandrovec

Problem is that pci_dev may be NULL - and it is NULL for example with
kernel I've just built, with amd IDE driver built as a module while with ide-disk built into the kernel.

Yes that was discussed extensively by Andi and me and finally fixed by Kiran's patch in 2.6.13-rc6.

A message that is not in the archive (sent off-list or not yet received): [email protected]

#3 lore raw
Quoted 6 lines · Petr Vandrovec
Quoted 2 lines

Yes that was discussed extensively by Andi and me and finally fixed by Kiran's patch in 2.6.13-rc6.

By which patch? I hit it with post-2.6.13-rc6, exactly 2.6.13-rc6 with checkin hash "commit 00dd1e433967872f3997a45d5adf35056fdf2f56". So if it is supposed to be fixed in 2.6.13-rc6, it is not.

Yes you are right there is one additional place where pcibus_to_node is used with the hwif that we did not cover. This better go into 2.6.13.

---
Fix ide-disk.c oops caused by hwif == NULL

1. Move hwif_to_node to ide.h

2. Use hwif_to_node in ide-disk.c

drivers/ide/ide-disk.c +1 -1
Index: linux-2.6/drivers/ide/ide-disk.c
===================================================================
@@ -1220,7 +1220,7 @@
 		goto failed;
 
 	g = alloc_disk_node(1 << PARTN_BITS,
-			pcibus_to_node(drive->hwif->pci_dev->bus));
+			hwif_to_node(drive->hwif));
 	if (!g)
 		goto out_free_idkp;
 
drivers/ide/ide-probe.c +0 -9
Index: linux-2.6/drivers/ide/ide-probe.c
===================================================================
@@ -960,15 +960,6 @@
 }
 #endif /* MAX_HWIFS > 1 */
 
-static inline int hwif_to_node(ide_hwif_t *hwif)
-{
-	if (hwif->pci_dev)
-		return pcibus_to_node(hwif->pci_dev->bus);
-	else
-		/* Add ways to determine the node of other busses here */
-		return -1;
-}
-
 /*
  * init request queue
  */
include/linux/ide.h +9 -0
Index: linux-2.6/include/linux/ide.h
===================================================================
@@ -1501,4 +1501,13 @@
 #define ide_id_has_flush_cache_ext(id)	\
 	(((id)->cfs_enable_2 & 0x2400) == 0x2400)
 
+static inline int hwif_to_node(ide_hwif_t *hwif)
+{
+	if (hwif->pci_dev)
+		return pcibus_to_node(hwif->pci_dev->bus);
+	else
+		/* Add ways to determine the node of other busses here */
+		return -1;
+}
+
 #endif /* _IDE_H */
#4 lore raw

Why have you removed me from cc:?

Quoted 11 lines · Christoph Lameter
Quoted 6 lines · Petr Vandrovec
Quoted 2 lines

Yes that was discussed extensively by Andi and me and finally fixed by Kiran's patch in 2.6.13-rc6.

By which patch? I hit it with post-2.6.13-rc6, exactly 2.6.13-rc6 with checkin hash "commit 00dd1e433967872f3997a45d5adf35056fdf2f56". So if it is supposed to be fixed in 2.6.13-rc6, it is not.

Yes you are right there is one additional place where pcibus_to_node is used with the hwif that we did not cover. This better go into 2.6.13.

Yes.

Quoted 2 lines

---
Fix ide-disk.c oops caused by hwif == NULL

This description is wrong, should be: hwif->pci_dev == NULL

Quoted 63 lines

1. Move hwif_to_node to ide.h

2. Use hwif_to_node in ide-disk.c

Index: linux-2.6/drivers/ide/ide-disk.c
===================================================================

drivers/ide/ide-disk.c 2005-08-09 19:55:03.000000000 -0700 renamed from drivers/ide/ide-disk.c 2005-07-27 18:29:17.000000000 -0700 +1 -1
@@ -1220,7 +1220,7 @@
                goto failed;
 
        g = alloc_disk_node(1 << PARTN_BITS,
-                       pcibus_to_node(drive->hwif->pci_dev->bus));
+                       hwif_to_node(drive->hwif));
        if (!g)
                goto out_free_idkp;
 

Index: linux-2.6/drivers/ide/ide-probe.c
===================================================================

drivers/ide/ide-probe.c 2005-08-09 19:46:50.000000000 -0700 renamed from drivers/ide/ide-probe.c 2005-08-04 15:47:15.000000000 -0700 +0 -9
@@ -960,15 +960,6 @@
 }
 #endif /* MAX_HWIFS > 1 */
 
-static inline int hwif_to_node(ide_hwif_t *hwif)
-{
-       if (hwif->pci_dev)
-               return pcibus_to_node(hwif->pci_dev->bus);
-       else
-               /* Add ways to determine the node of other busses here */
-               return -1;
-}
-
 /*
  * init request queue
  */

Index: linux-2.6/include/linux/ide.h
===================================================================

include/linux/ide.h 2005-08-09 19:47:14.000000000 -0700 renamed from include/linux/ide.h 2005-07-27 18:29:23.000000000 -0700 +9 -1
@@ -1501,4 +1501,13 @@
 #define ide_id_has_flush_cache_ext(id) \
        (((id)->cfs_enable_2 & 0x2400) == 0x2400)
 
+static inline int hwif_to_node(ide_hwif_t *hwif)
+{
+       if (hwif->pci_dev)
+               return pcibus_to_node(hwif->pci_dev->bus);
+       else
+               /* Add ways to determine the node of other busses here */
+               return -1;
+}
+
 #endif /* _IDE_H */
-

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

#5 lore raw
In reply to Christoph Lameter
Quoted 2 lines · Christoph Lameter

Yes you are right there is one additional place where pcibus_to_node is used with the hwif that we did not cover. This better go into 2.6.13.

drive->hwif is not permitted to be NULL. Please work back and fix the actual bug.

Quoted 1 line
#6 lore raw
Quoted 6 lines · Alan Cox
Quoted 2 lines · Christoph Lameter

Yes you are right there is one additional place where pcibus_to_node is used with the hwif that we did not cover. This better go into 2.6.13.

drive->hwif is not permitted to be NULL. Please work back and fix the actual bug.

The actual bug was fixed. The description was wrong. We are not checking hwif but hwif->pci_dev.