[PATCH v6 9/9] x86: Pass memory range via E820 for kdump

[WANG Chao]

On 04/21/14 at 09:01am, Simon Horman wrote:
> On Thu, Apr 17, 2014 at 01:58:17PM +0800, Dave Young wrote:
> > On 04/17/14 at 01:48pm, WANG Chao wrote:
> > > On 04/17/14 at 01:29pm, Dave Young wrote:
> > > > On 04/14/14 at 10:55pm, WANG Chao wrote:
> > > > > command line size is restricted by kernel, sometimes memmap=exactmap has
> > > > > too many memory ranges to pass to cmdline. And also memmap=exactmap and
> > > > > kASLR doesn't work together.
> > > > > 
> > > > > A better approach, to pass the memory ranges for crash kernel to boot
> > > > > into, is filling the memory ranges into E820.
> > > > > 
> > > > > boot_params only got 128 slots for E820 map to fit in, when the number of
> > > > > memory map exceeds 128, use setup_data to pass the rest as extended E820
> > > > > memory map.
> > > > > 
> > > > > kexec boot could also benefit from setup_data in case E820 memory map
> > > > > exceeds 128.
> > > > > 
> > > > > Now this new approach becomes default instead of memmap=exactmap.
> > > > > saved_max_pfn users can specify --pass-memmap-cmdline to use the
> > > > > exactmap approach.
> > > > > 
> > > > > Signed-off-by: WANG Chao <chaowang at redhat.com>
> > > > > Tested-by: Linn Crosetto <linn at hp.com>
> > > > > Reviewed-by: Linn Crosetto <linn at hp.com>
> > > > > ---
> > > > >  kexec/arch/i386/crashdump-x86.c   |   6 +-
> > > > >  kexec/arch/i386/x86-linux-setup.c | 149 +++++++++++++++++++++++++-------------
> > > > >  kexec/arch/i386/x86-linux-setup.h |   1 +
> > > > >  3 files changed, 103 insertions(+), 53 deletions(-)
> > > > > 
> > > > > diff --git a/kexec/arch/i386/crashdump-x86.c b/kexec/arch/i386/crashdump-x86.c
> > > > > index 7b618a6..4a1491b 100644
> > > > > --- a/kexec/arch/i386/crashdump-x86.c
> > > > > +++ b/kexec/arch/i386/crashdump-x86.c
> > > > > @@ -979,7 +979,8 @@ int load_crashdump_segments(struct kexec_info *info, char* mod_cmdline,
> > > > >  	dbgprintf("Created elf header segment at 0x%lx\n", elfcorehdr);
> > > > >  	if (delete_memmap(memmap_p, &nr_memmap, elfcorehdr, memsz) < 0)
> > > > >  		return -1;
> > > > > -	cmdline_add_memmap(mod_cmdline, memmap_p);
> > > > > +	if (arch_options.pass_memmap_cmdline)
> > > > > +		cmdline_add_memmap(mod_cmdline, memmap_p);
> > > > >  	if (!bzImage_support_efi_boot)
> > > > >  		cmdline_add_efi(mod_cmdline);
> > > > >  	cmdline_add_elfcorehdr(mod_cmdline, elfcorehdr);
> > > > > @@ -995,7 +996,8 @@ int load_crashdump_segments(struct kexec_info *info, char* mod_cmdline,
> > > > >  		type = mem_range[i].type;
> > > > >  		size = end - start + 1;
> > > > >  		add_memmap(memmap_p, &nr_memmap, start, size, type);
> > > > > -		cmdline_add_memmap_acpi(mod_cmdline, start, end);
> > > > > +		if (arch_options.pass_memmap_cmdline)
> > > > > +			cmdline_add_memmap_acpi(mod_cmdline, start, end);
> > > > 
> > > > Seems memmap_p contains the acpi ranges as well, so cmdline_add_memmap_acpi is
> > > > not necessary anymore, just improve cmdline_add_memmap to add both RAM and ACPI
> > > > ranges is enough.
> > > 
> > > I can do that. But is it what the patchset is really about ...
> > > 
> > > I'm not keen about doing too much cleanup in this series now since it's
> > > already v6. I really want to get this in as early as possible to
> > > cope with calgary iommu change in the kernel.
> > > 
> > > I prefer to first get this patch in if there's no problem in it and then
> > > look back and think about how we can clean up the code block which have
> > > been there for historical reason.
> > 
> > I think the cleanup is worth, but if you want to do it later I'm fine.
> > So should better leave the patch 5/9 to later cleanup as well.
> > 
> > Thus except for 5/9, for other patches:
> > Acked-by: Dave Young <dyoung at redhat.com>

Hi, Simon

Later, after a discussion with Dave Young, he is ok with 5/9. Here's his
ACK for the whole patch series:

http://lists.infradead.org/pipermail/kexec/2014-April/011615.html

I'll explain why 5/9 is necessary below.

> 
> This sounds fine to me.

Patch 5/9 increased memmap_p from a small array to a large one. That
patch is necessary.

Originally memmap_p was used to store the RANGE_RAM only for 2nd kernel
and that was fine for memmap_p to be small, since we only have several
memory range of RANGE_RAM to be passed to 2nd kernel.

However now memmap_p will be used to store all types of memory ranges to
pass to 2nd kernel, which means not only RANGE_RAM but also RANGE_ACPI
and RANGE_ACPI_NVS (probably RANGE_RESERVED in the future). That small
memmap_p is not likely to be enough to contain all the memory ranges. So
that's why we should increase CRASH_MAX_MEMMAP_NR to about 1024 to adapt
the change of memmap_p will contains all type of memory ranges.

Thanks
WANG Chao