Re: [PATCH v12 12/16] arm64: kexec_file: add crash dump support

From: AKASHI Takahiro
Date: Fri Jul 27 2018 - 02:59:03 EST


On Thu, Jul 26, 2018 at 02:36:58PM +0100, James Morse wrote:
> Hi Akashi,
>
> On 24/07/18 07:57, AKASHI Takahiro wrote:
> > Enabling crash dump (kdump) includes
> > * prepare contents of ELF header of a core dump file, /proc/vmcore,
> > using crash_prepare_elf64_headers(), and
> > * add two device tree properties, "linux,usable-memory-range" and
> > "linux,elfcorehdr", which represent respectively a memory range
> > to be used by crash dump kernel and the header's location
>
> > diff --git a/arch/arm64/include/asm/kexec.h b/arch/arm64/include/asm/kexec.h
> > index 5d102a1054b3..1b2c27026ae0 100644
> > --- a/arch/arm64/include/asm/kexec.h
> > +++ b/arch/arm64/include/asm/kexec.h
> > @@ -97,8 +97,12 @@ static inline void crash_post_resume(void) {}
> > #define ARCH_HAS_KIMAGE_ARCH
> >
> > struct kimage_arch {
>
> > - void *dtb_buf;
> > + void *dtb;
>
> This change should be in an earlier patch, otherwise this series doesn't build
> during bisect.

Will fix.

> With the build-issues fixed:
> Reviewed-by: James Morse <james.morse@xxxxxxx>
>
> Some boring Nits:
>
> > unsigned long dtb_mem;
> > + /* Core ELF header buffer */
> > + void *elf_headers;
> > + unsigned long elf_headers_mem;
> > + unsigned long elf_headers_sz;
> > };
> > diff --git a/arch/arm64/kernel/machine_kexec_file.c b/arch/arm64/kernel/machine_kexec_file.c
> > index b8297f10e2ef..7356da5a53d5 100644
> > --- a/arch/arm64/kernel/machine_kexec_file.c
> > +++ b/arch/arm64/kernel/machine_kexec_file.c
> > @@ -38,12 +44,30 @@ static int setup_dtb(struct kimage *image,
> > void **dtb_buf, unsigned long *dtb_buf_len)
> > {
> > void *buf = NULL;
> > - size_t buf_size;
> > + size_t buf_size, range_size;
> > int nodeoffset;
> > int ret;
> >
> > /* duplicate dt blob */
> > buf_size = fdt_totalsize(initial_boot_params);
> > + range_size = of_fdt_reg_cells_size();
> > +
> > + if (image->type == KEXEC_TYPE_CRASH) {
> > + buf_size += fdt_prop_len("linux,elfcorehdr", range_size);
> > + buf_size += fdt_prop_len("linux,usable-memory-range",
> > + range_size);
> > + }
>
> Nit: it would be better if these strings were defined in a header file somewhere
> so we don't risk a typo if this gets refactored.

Nit??
Well, I do understand your concern, but it's a bit headache.
If we handle them in such a way, we may want to handle others, such as
linux,initrd-start/end, chosen and bootargs, as well in this file.
They are not kexec specific and should go into a common header. This will
end up propagating similar changes to other no-kexec-related occurrences
under drivers/of.

So I want to keep the following def's local in this file for easy
maintenance.
#define FDT_PSTR_KEXEC_ELFHDR "linux,elfcorehdr"
#define FDT_PSTR_MEM_RANGE "linux,usable-memory-range"
#define FDT_PSTR_INITRD_ST "linux,initrd-start"
#define FDT_PSTR_INITRD_END "linux,initrd-end"
#define FDT_PSTR_BOOTARGS "bootargs"
#define FDT_PSTR_KASLR_SEED "kaslr-seed"

>
>
> > @@ -129,6 +170,43 @@ static int setup_dtb(struct kimage *image,
> > return ret;
> > }
> >
> > +static int prepare_elf_headers(void **addr, unsigned long *sz)
> > +{
> > + struct crash_mem *cmem;
> > + unsigned int nr_ranges;
> > + int ret;
> > + u64 i;
> > + phys_addr_t start, end;
> > +
> > + nr_ranges = 1; /* for exclusion of crashkernel region */
>
> > + for_each_mem_range(i, &memblock.memory, NULL, NUMA_NO_NODE, 0,
> > + &start, &end, NULL)
>
> Nit: MEMBLOCK_NONE

OK

-Takahiro AKASHI


>
> Thanks,
>
> James