2008-10-12 08:51:18

by Petr Vandrovec

Subject: [PATCH] Fix early panic on amd64 due to typo in supported CPU section

It seems that I forgot to CC LKML on that one...
Petr

----- Forwarded message from Petr Vandrovec <[email protected]> -----

Date: Sun, 12 Oct 2008 10:01:26 +0200
From: Petr Vandrovec <[email protected]>
To: [email protected]
Cc: [email protected], [email protected], [email protected]
Subject: [PATCH] Fix early panic on amd64 due to typo in supported CPU section

Do not crash when enumerating supported CPU architectures

SECURITY_INIT somehow ended up in x86_cpu_dev.init section. That caused printk
in code which prints supported architectures to hit #GP due to non-canonical
address being used.

Signed-off-by: Petr Vandrovec <[email protected]>

diff --git a/arch/x86/kernel/vmlinux_64.lds.S b/arch/x86/kernel/vmlinux_64.lds.S
index 201e81a..46e0544 100644
--- a/arch/x86/kernel/vmlinux_64.lds.S
+++ b/arch/x86/kernel/vmlinux_64.lds.S
@@ -172,8 +172,8 @@ SECTIONS
.x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) {
*(.x86_cpu_dev.init)
}
- SECURITY_INIT
__x86_cpu_dev_end = .;
+ SECURITY_INIT

. = ALIGN(8);
.parainstructions : AT(ADDR(.parainstructions) - LOAD_OFFSET) {

----- End forwarded message -----

2008-10-12 09:22:56

by Ingo Molnar

Subject: Re: [PATCH] Fix early panic on amd64 due to typo in supported CPU section

* Petr Vandrovec <[email protected]> wrote:

> Do not crash when enumerating supported CPU architectures
>
> SECURITY_INIT somehow ended up in x86_cpu_dev.init section. That caused printk
> in code which prints supported architectures to hit #GP due to non-canonical
> address being used.
>
> Signed-off-by: Petr Vandrovec <[email protected]>

thanks Petr, we had this bug fixed later in the patches but lets pick
your fix right now before merging anything else. Applied it to
tip/x86/urgent.

Linus,

Please pull the latest x86-fixes-for-linus git tree from:

git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git x86-fixes-for-linus

Thanks,

Ingo

------------------>
Alan Cox (1):
x86, early_ioremap: fix fencepost error

David Rientjes (1):
x86: avoid dereferencing beyond stack + THREAD_SIZE

Petr Vandrovec (1):
x86: fix early panic on amd64 due to typo in supported CPU section


arch/x86/kernel/process_64.c | 4 ++--
arch/x86/kernel/vmlinux_64.lds.S | 2 +-
arch/x86/mm/ioremap.c | 2 +-
3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 2a8ccb9..b6b508e 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -754,12 +754,12 @@ unsigned long get_wchan(struct task_struct *p)
if (!p || p == current || p->state == TASK_RUNNING)
return 0;
stack = (unsigned long)task_stack_page(p);
- if (p->thread.sp < stack || p->thread.sp > stack+THREAD_SIZE)
+ if (p->thread.sp < stack || p->thread.sp >= stack+THREAD_SIZE)
return 0;
fp = *(u64 *)(p->thread.sp);
do {
if (fp < (unsigned long)stack ||
- fp > (unsigned long)stack+THREAD_SIZE)
+ fp >= (unsigned long)stack+THREAD_SIZE)
return 0;
ip = *(u64 *)(fp+8);
if (!in_sched_functions(ip))
diff --git a/arch/x86/kernel/vmlinux_64.lds.S b/arch/x86/kernel/vmlinux_64.lds.S
index 201e81a..46e0544 100644
--- a/arch/x86/kernel/vmlinux_64.lds.S
+++ b/arch/x86/kernel/vmlinux_64.lds.S
@@ -172,8 +172,8 @@ SECTIONS
.x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) {
*(.x86_cpu_dev.init)
}
- SECURITY_INIT
__x86_cpu_dev_end = .;
+ SECURITY_INIT

. = ALIGN(8);
.parainstructions : AT(ADDR(.parainstructions) - LOAD_OFFSET) {
diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c
index 6ab3196..10b5230 100644
--- a/arch/x86/mm/ioremap.c
+++ b/arch/x86/mm/ioremap.c
@@ -614,7 +614,7 @@ void __init *early_ioremap(unsigned long phys_addr, unsigned long size)
*/
offset = phys_addr & ~PAGE_MASK;
phys_addr &= PAGE_MASK;
- size = PAGE_ALIGN(last_addr) - phys_addr;
+ size = PAGE_ALIGN(last_addr + 1) - phys_addr;

/*
* Mappings have to fit in the FIX_BTMAP area.