Patchwork crash with /proc/iomem on sparc64

login
register
mail settings
Submitter Mikulas Patocka
Date March 17, 2009, 12:03 p.m.
Message ID <Pine.LNX.4.64.0903170758060.24107@hs20-bc2-1.build.redhat.com>
Download mbox | patch
Permalink /patch/24557/
State Accepted
Delegated to: David Miller
Headers show

Comments

Mikulas Patocka - March 17, 2009, 12:03 p.m.
Hi

When you compile kernel on Sparc64 with heap memory checking and type
"cat /proc/iomem", you get a crash, because pointers in struct resource 
are uninitialized.

Most code fills struct resource with zeros, so I assume that it is 
responsibility of the caller of request_resource to initialized it, not 
the responsibility of request_resource functuion.

After 2.6.29 is out, there could be a check for uninitialized fields added 
to request_resource to avoid crashes like this.

Mikulas

---
 arch/sparc/kernel/pci_common.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Mikulas Patocka - March 17, 2009, 12:08 p.m.
On Tue, 17 Mar 2009, Mikulas Patocka wrote:

> Hi
> 
> When you compile kernel on Sparc64 with heap memory checking and type
> "cat /proc/iomem", you get a crash, because pointers in struct resource 
> are uninitialized.
> 
> Most code fills struct resource with zeros, so I assume that it is 
> responsibility of the caller of request_resource to initialized it, not 
> the responsibility of request_resource functuion.
> 
> After 2.6.29 is out, there could be a check for uninitialized fields added 
> to request_resource to avoid crashes like this.
> 
> Mikulas

I forgot to add:

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>

> ---
>  arch/sparc/kernel/pci_common.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> Index: linux-2.6.29-rc8-devel/arch/sparc/kernel/pci_common.c
> ===================================================================
> --- linux-2.6.29-rc8-devel.orig/arch/sparc/kernel/pci_common.c	2009-03-17 12:49:51.000000000 +0100
> +++ linux-2.6.29-rc8-devel/arch/sparc/kernel/pci_common.c	2009-03-17 12:50:16.000000000 +0100
> @@ -368,7 +368,7 @@ static void pci_register_iommu_region(st
>  	const u32 *vdma = of_get_property(pbm->op->node, "virtual-dma", NULL);
>  
>  	if (vdma) {
> -		struct resource *rp = kmalloc(sizeof(*rp), GFP_KERNEL);
> +		struct resource *rp = kzalloc(sizeof(*rp), GFP_KERNEL);
>  
>  		if (!rp) {
>  			prom_printf("Cannot allocate IOMMU resource.\n");
> 
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Mikulas Patocka - March 17, 2009, 1:35 p.m.
On Tue, 17 Mar 2009, Mikulas Patocka wrote:

> 
> 
> On Tue, 17 Mar 2009, Mikulas Patocka wrote:
> 
> > Hi
> > 
> > When you compile kernel on Sparc64 with heap memory checking and type
> > "cat /proc/iomem", you get a crash, because pointers in struct resource 
> > are uninitialized.
> > 
> > Most code fills struct resource with zeros, so I assume that it is 
> > responsibility of the caller of request_resource to initialized it, not 
> > the responsibility of request_resource functuion.
> > 
> > After 2.6.29 is out, there could be a check for uninitialized fields added 
> > to request_resource to avoid crashes like this.
> > 
> > Mikulas
> 
> I forgot to add:
> 
> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>

... and

Cc: stable@kernel.org

because it applies to old kernels as well, even 2.6.26 crashes.

Mikulas

> > ---
> >  arch/sparc/kernel/pci_common.c |    2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > Index: linux-2.6.29-rc8-devel/arch/sparc/kernel/pci_common.c
> > ===================================================================
> > --- linux-2.6.29-rc8-devel.orig/arch/sparc/kernel/pci_common.c	2009-03-17 12:49:51.000000000 +0100
> > +++ linux-2.6.29-rc8-devel/arch/sparc/kernel/pci_common.c	2009-03-17 12:50:16.000000000 +0100
> > @@ -368,7 +368,7 @@ static void pci_register_iommu_region(st
> >  	const u32 *vdma = of_get_property(pbm->op->node, "virtual-dma", NULL);
> >  
> >  	if (vdma) {
> > -		struct resource *rp = kmalloc(sizeof(*rp), GFP_KERNEL);
> > +		struct resource *rp = kzalloc(sizeof(*rp), GFP_KERNEL);
> >  
> >  		if (!rp) {
> >  			prom_printf("Cannot allocate IOMMU resource.\n");
> > 
> 
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Miller - March 17, 2009, 7:15 p.m.
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 17 Mar 2009 09:35:23 -0400 (EDT)

> Cc: stable@kernel.org
> 
> because it applies to old kernels as well, even 2.6.26 crashes.

-stable will be sent a copy by me when it hits Linus's tree, and no
sooner.
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Miller - March 19, 2009, 6:53 a.m.
From: Mikulas Patocka <mpatocka@redhat.com>
Date: Tue, 17 Mar 2009 08:08:08 -0400 (EDT)

> 
> 
> On Tue, 17 Mar 2009, Mikulas Patocka wrote:
> 
> > Hi
> > 
> > When you compile kernel on Sparc64 with heap memory checking and type
> > "cat /proc/iomem", you get a crash, because pointers in struct resource 
> > are uninitialized.
> > 
> > Most code fills struct resource with zeros, so I assume that it is 
> > responsibility of the caller of request_resource to initialized it, not 
> > the responsibility of request_resource functuion.
> > 
> > After 2.6.29 is out, there could be a check for uninitialized fields added 
> > to request_resource to avoid crashes like this.
> > 
> > Mikulas
> 
> I forgot to add:
> 
> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>

Patch applied and queued up for -stable, thanks a lot.
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

Index: linux-2.6.29-rc8-devel/arch/sparc/kernel/pci_common.c
===================================================================
--- linux-2.6.29-rc8-devel.orig/arch/sparc/kernel/pci_common.c	2009-03-17 12:49:51.000000000 +0100
+++ linux-2.6.29-rc8-devel/arch/sparc/kernel/pci_common.c	2009-03-17 12:50:16.000000000 +0100
@@ -368,7 +368,7 @@  static void pci_register_iommu_region(st
 	const u32 *vdma = of_get_property(pbm->op->node, "virtual-dma", NULL);
 
 	if (vdma) {
-		struct resource *rp = kmalloc(sizeof(*rp), GFP_KERNEL);
+		struct resource *rp = kzalloc(sizeof(*rp), GFP_KERNEL);
 
 		if (!rp) {
 			prom_printf("Cannot allocate IOMMU resource.\n");