[PATCH][v2] x86/process: Return early if TIF_IO_BITMAP is set but no IO bitmap is assigned

[PATCH][v2] x86/process: Return early if TIF_IO_BITMAP is set but no IO bitmap is assigned

[lirongqing]

From: Li RongQing <lirongqing@baidu.com>

Outside the IOPL emulation path, the IO bitmap is always expected
to be allocated when TIF_IO_BITMAP is set. The paranoid WARN_ON_ONCE()
handles the case where the flag and the pointer got out of sync.
In this rare scenario, return early instead of continuing and
dereferencing a NULL pointer.

[Sohil Mehta: rephrased commit message]

Reviewed-by: Sohil Mehta <sohil.mehta@intel.com>
Signed-off-by: Li RongQing <lirongqing@baidu.com>
---
Diff with v1: rephrased commit message

 arch/x86/kernel/process.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 4c718f8..d5cd217 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -486,6 +486,7 @@ void native_tss_update_io_bitmap(void)
 		if (WARN_ON_ONCE(!iobm)) {
 			clear_thread_flag(TIF_IO_BITMAP);
 			native_tss_invalidate_io_bitmap();
+			return;
 		}
 
 		/*
-- 
2.9.4

Re: [PATCH][v2] x86/process: Return early if TIF_IO_BITMAP is set but no IO bitmap is assigned

[Sohil Mehta]

On 6/11/2026 10:49 PM, lirongqing wrote:
> From: Li RongQing <lirongqing@baidu.com>
> 
> Outside the IOPL emulation path, the IO bitmap is always expected
> to be allocated when TIF_IO_BITMAP is set. The paranoid WARN_ON_ONCE()
> handles the case where the flag and the pointer got out of sync.
> In this rare scenario, return early instead of continuing and
> dereferencing a NULL pointer.
> 
> [Sohil Mehta: rephrased commit message]
> 

Never add handler comments for others. Such comments are only expected
to be added when someone who is handling your patch modifies it but
keeps the original authorship. You shouldn't add handler comments when
you are changing the patch as an author yourself.

> Reviewed-by: Sohil Mehta <sohil.mehta@intel.com>
> Signed-off-by: Li RongQing <lirongqing@baidu.com>
> ---

答复: [外部邮件] Re: [PATCH][v2] x86/process: Return early if TIF_IO_BITMAP is set but no IO bitmap is assigned

[Li,Rongqing]

> On 6/11/2026 10:49 PM, lirongqing wrote:
> > From: Li RongQing <lirongqing@baidu.com>
> >
> > Outside the IOPL emulation path, the IO bitmap is always expected to
> > be allocated when TIF_IO_BITMAP is set. The paranoid WARN_ON_ONCE()
> > handles the case where the flag and the pointer got out of sync.
> > In this rare scenario, return early instead of continuing and
> > dereferencing a NULL pointer.
> >
> > [Sohil Mehta: rephrased commit message]
> >
> 
> Never add handler comments for others. Such comments are only expected
> to be added when someone who is handling your patch modifies it but keeps
> the original authorship. You shouldn't add handler comments when you are
> changing the patch as an author yourself.
> 

Ok, I will remove it in next version

[Li,Rongqing] 

Re: 答复: [外部邮件] Re: [PATCH][v2] x86/process: Return early if TIF_IO_BITMAP is set but no IO bitmap is assigned

[Sohil Mehta]

Upon rereading, the title seems very long. Maybe shorten it to:

x86/ioperm: Prevent NULL dereference on missing IO bitmap

On 6/12/2026 8:01 PM, Li,Rongqing wrote:

>>> [Sohil Mehta: rephrased commit message]
>>>

Such comments are usually added below the '---' line if needed.

> Ok, I will remove it in next version
> 
> [Li,Rongqing] 
> 
>