Bug 201637 - hangup sinse 'x86/mm: Expand static page table for fixmap space' with CONFIG_RANDOMIZE_BASE enabled
Summary: hangup sinse 'x86/mm: Expand static page table for fixmap space' with CONFIG...
Status: NEW
Alias: None
Product: Memory Management
Classification: Unclassified
Component: Other (show other bugs)
Hardware: All Linux
: P1 high
Assignee: Andrew Morton
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-11-08 13:48 UTC by Anton Boyarshinov
Modified: 2018-11-09 02:58 UTC (History)
1 user (show)

See Also:
Kernel Version: 4.9.133
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description Anton Boyarshinov 2018-11-08 13:48:25 UTC
I have early kernel hungup on x86_64 since 4.9.133 and i've finaly bisected it.

This commit is incompatible with CONFIG_RANDOMIZE_BASE enabled and cause immediate silent hangup during boot. 

commit 3a8304b7ad2e291777e8499e39390145d932a2fd
Author: Feng Tang <feng.tang@intel.com>
Date:   Thu Sep 20 10:58:28 2018 +0800

    x86/mm: Expand static page table for fixmap space
    
    commit 05ab1d8a4b36ee912b7087c6da127439ed0a903e upstream.
    
    We met a kernel panic when enabling earlycon, which is due to the fixmap
    address of earlycon is not statically setup.
Comment 1 Andrew Morton 2018-11-08 22:20:55 UTC
(switched to email.  Please respond via emailed reply-to-all, not via the
bugzilla web interface).

On Thu, 08 Nov 2018 13:48:25 +0000 bugzilla-daemon@bugzilla.kernel.org wrote:

> https://bugzilla.kernel.org/show_bug.cgi?id=201637
> 
>             Bug ID: 201637
>            Summary: hangup sinse  'x86/mm: Expand static page table for
>                     fixmap space' with CONFIG_RANDOMIZE_BASE enabled
>            Product: Memory Management
>            Version: 2.5
>     Kernel Version: 4.9.133
>           Hardware: All
>                 OS: Linux
>               Tree: Mainline
>             Status: NEW
>           Severity: high
>           Priority: P1
>          Component: Other
>           Assignee: akpm@linux-foundation.org
>           Reporter: boyarsh@altlinux.org
>         Regression: No
> 
> I have early kernel hungup on x86_64 since 4.9.133 and i've finaly bisected
> it.
> 
> This commit is incompatible with CONFIG_RANDOMIZE_BASE enabled and cause
> immediate silent hangup during boot. 
> 
> commit 3a8304b7ad2e291777e8499e39390145d932a2fd
> Author: Feng Tang <feng.tang@intel.com>
> Date:   Thu Sep 20 10:58:28 2018 +0800
> 
>     x86/mm: Expand static page table for fixmap space
> 
>     commit 05ab1d8a4b36ee912b7087c6da127439ed0a903e upstream.
> 
>     We met a kernel panic when enabling earlycon, which is due to the fixmap
>     address of earlycon is not statically setup.
> 
> -- 
> You are receiving this mail because:
> You are the assignee for the bug.
Comment 2 Feng Tang 2018-11-09 01:32:25 UTC
Hi All,

On Thu, Nov 08, 2018 at 02:20:52PM -0800, Andrew Morton wrote:
> 
> (switched to email.  Please respond via emailed reply-to-all, not via the
> bugzilla web interface).
> 
> On Thu, 08 Nov 2018 13:48:25 +0000 bugzilla-daemon@bugzilla.kernel.org wrote:
> 
> > https://bugzilla.kernel.org/show_bug.cgi?id=201637
> > 
> >             Bug ID: 201637
> >            Summary: hangup sinse  'x86/mm: Expand static page table for
> >                     fixmap space' with CONFIG_RANDOMIZE_BASE enabled
> >            Product: Memory Management
> >            Version: 2.5
> >     Kernel Version: 4.9.133
> >           Hardware: All
> >                 OS: Linux
> >               Tree: Mainline
> >             Status: NEW
> >           Severity: high
> >           Priority: P1
> >          Component: Other
> >           Assignee: akpm@linux-foundation.org
> >           Reporter: boyarsh@altlinux.org
> >         Regression: No
> > 
> > I have early kernel hungup on x86_64 since 4.9.133 and i've finaly bisected
> it.
> > 
> > This commit is incompatible with CONFIG_RANDOMIZE_BASE enabled and cause
> > immediate silent hangup during boot. 

Thanks for reporting. The patch revert was queued already. see
https://lkml.org/lkml/2018/11/2/453

And you are right, the problem is related with the kernel address randomization,
the original patch in mainline covers it, and the problem happens during my manual
backporting to 4.9, where I didn't fine head64.c and wrongly assumed KASLR was
not in 4.9 :(   sorry for the trouble. 

Thanks,
Feng

Note You need to log in before you can comment on or make changes to this bug.