Re: [PATCH] RISC-V: Add address map dumper

From: Logan Gunthorpe
Date: Mon Nov 11 2019 - 13:14:45 EST




On 2019-11-10 10:27 p.m., Yash Shah wrote:
> Add support for dumping the kernel address space layout to the console.
> User can enable CONFIG_DEBUG_VM_LAYOUT to dump the virtual memory region
> into dmesg buffer during boot-up.

Cool, I'd find this useful. Though, is there any reason we don't do this
more generally for all architectures?

> Signed-off-by: Yash Shah <yash.shah@xxxxxxxxxx>
> ---
> This patch is based on Linux 5.4-rc6 and tested on SiFive HiFive Unleashed
> board.
> ---
> arch/riscv/Kconfig.debug | 9 +++++++++
> arch/riscv/mm/init.c | 30 ++++++++++++++++++++++++++++++
> 2 files changed, 39 insertions(+)
>
> diff --git a/arch/riscv/Kconfig.debug b/arch/riscv/Kconfig.debug
> index e69de29..cdedfd3 100644
> --- a/arch/riscv/Kconfig.debug
> +++ b/arch/riscv/Kconfig.debug
> @@ -0,0 +1,9 @@
> +config DEBUG_VM_LAYOUT
> + bool "Print virtual memory layout on boot up"
> + depends on DEBUG_KERNEL
> + help
> + Say Y here if you want to dump the kernel virtual memory layout to
> + dmesg log on boot up. This information is only useful for kernel
> + developers who are working in architecture specific areas of the
> + kernel. It is probably not a good idea to enable this feature in a
> + production kernel.
> diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
> index 79cfb35..fcb8144 100644
> --- a/arch/riscv/mm/init.c
> +++ b/arch/riscv/mm/init.c
> @@ -55,6 +55,36 @@ void __init mem_init(void)
> memblock_free_all();
>
> mem_init_print_info(NULL);
> +#ifdef CONFIG_DEBUG_VM_LAYOUT

Generally, it's best to avoid #ifdefs inside functions, it's even
counter-indicated in the style guide[1].

> +#define MLK(b, t) b, t, (((t) - (b)) >> 10)
> +#define MLM(b, t) b, t, (((t) - (b)) >> 20)
> +#define MLK_ROUNDUP(b, t) b, t, DIV_ROUND_UP(((t) - (b)), SZ_1K)

I personally find these inline defines rather ugly. Maybe it would be
better to have a helper function that prints a single line. Also seems
like MLK and MLK_ROUNDUP could be the same assuming the entries in MLK
are aligned...

> +
> + pr_notice("Virtual kernel memory layout:\n"
> + " fixmap : 0x%08lx - 0x%08lx (%4ld kB)\n"
> + " vmemmap : 0x%08lx - 0x%08lx (%4ld MB)\n"
> + " vmalloc : 0x%08lx - 0x%08lx (%4ld MB)\n"
> + " lowmem : 0x%08lx - 0x%08lx (%4ld MB)\n"
> + " .init : 0x%px - 0x%px (%4td kB)\n"
> + " .text : 0x%px - 0x%px (%4td kB)\n"
> + " .data : 0x%px - 0x%px (%4td kB)\n"
> + " .bss : 0x%px - 0x%px (%4td kB)\n",
> +
> + MLK(FIXADDR_START, FIXADDR_TOP),
> + MLM(VMEMMAP_START, VMEMMAP_END),
> + MLM(VMALLOC_START, VMALLOC_END),
> + MLM(PAGE_OFFSET, (unsigned long)high_memory),
> +
> + MLK_ROUNDUP(__init_begin, __init_end),
> + MLK_ROUNDUP(_text, _etext),
> + MLK_ROUNDUP(_sdata, _edata),
> + MLK_ROUNDUP(__bss_start, __bss_stop));
> +
> +#undef MLK
> +#undef MLM
> +#undef MLK_ROUNDUP
> +#endif
> }
>
> #ifdef CONFIG_BLK_DEV_INITRD

Thanks,

Logan

[1]
https://www.kernel.org/doc/html/latest/process/coding-style.html#conditional-compilation