[PATCH] Fix kdump Crash Kernel boot memory reservation for NUMA machines

[PATCH] Fix kdump Crash Kernel boot memory reservation for NUMA machines

[Amul Shah]

This patch will fix a boot memory reservation bug that trashes memory on
the ES7000 when loading the kdump crash kernel.

The code in arch/x86_64/kernel/setup.c to reserve boot memory for the 
crash kernel uses the non-numa aware "reserve_bootmem" function instead 
of the NUMA aware "reserve_bootmem_generic".  I checked to make sure 
that no other function was using "reserve_bootmem" and found none, 
except the ones that had NUMA ifdef'ed out.

I have tested this patch only on an ES7000 with NUMA on and off (numa=off)
in a single (non-NUMA) and multi-cell (NUMA) configurations.

Signed-off-by: Amul Shah <amul.shah@unisys.com>


---
--- linux-2.6.16.18-1.8/arch/x86_64/kernel/setup.c      2006-06-06 12:07:42.000000000 -0400
+++ linux-2.6.16.18-1.8-az/arch/x86_64/kernel/setup.c   2006-06-21 17:06:04.000000000 -0400
@@ -715,7 +715,7 @@ void __init setup_arch(char **cmdline_p)
 #endif
 #ifdef CONFIG_KEXEC
        if (crashk_res.start != crashk_res.end) {
-               reserve_bootmem(crashk_res.start,
+               reserve_bootmem_generic(crashk_res.start,
                        crashk_res.end - crashk_res.start + 1);
        }
 #endif

Re: [Fastboot] [PATCH] Fix kdump Crash Kernel boot memory reservation for NUMA machines

[Vivek Goyal]

On Fri, Jun 23, 2006 at 02:57:17PM -0400, Amul Shah wrote:
> This patch will fix a boot memory reservation bug that trashes memory on
> the ES7000 when loading the kdump crash kernel.
> 
> The code in arch/x86_64/kernel/setup.c to reserve boot memory for the 
> crash kernel uses the non-numa aware "reserve_bootmem" function instead 
> of the NUMA aware "reserve_bootmem_generic".  I checked to make sure 
> that no other function was using "reserve_bootmem" and found none, 
> except the ones that had NUMA ifdef'ed out.
> 
> I have tested this patch only on an ES7000 with NUMA on and off (numa=off)
> in a single (non-NUMA) and multi-cell (NUMA) configurations.
> 
> Signed-off-by: Amul Shah <amul.shah@unisys.com>
> 
> 
> ---
> --- linux-2.6.16.18-1.8/arch/x86_64/kernel/setup.c      2006-06-06 12:07:42.000000000 -0400
> +++ linux-2.6.16.18-1.8-az/arch/x86_64/kernel/setup.c   2006-06-21 17:06:04.000000000 -0400
> @@ -715,7 +715,7 @@ void __init setup_arch(char **cmdline_p)
>  #endif
>  #ifdef CONFIG_KEXEC
>         if (crashk_res.start != crashk_res.end) {
> -               reserve_bootmem(crashk_res.start,
> +               reserve_bootmem_generic(crashk_res.start,
>                         crashk_res.end - crashk_res.start + 1);
>         }

Looks good to me. I know of a 64w NUMA machine test results and kdump
was successful. Not sure why did not we see the problem there. But
anyway this is logical.

Thanks
Vivek