Re: [PATCH v2 3/5] riscv: Add support for kexec handover (KHO)

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] riscv: kho: bootmem reservation fails when DRAM starts at or above 4 GiB
--

commit 25f3cc12aa8edcc1fbb1363f837f61e462a57ad6
Author: Chen Pei <cp0613@linux.alibaba.com>

riscv: Add support for kexec handover (KHO)

This commit enables kexec handover (KHO) support on 64-bit RISC-V systems with
MMU enabled by defining ARCH_SUPPORTS_KEXEC_HANDOVER in arch/riscv/Kconfig.

> diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
> index 435da07132437..e11463a923a32 100644
> --- a/arch/riscv/Kconfig
> +++ b/arch/riscv/Kconfig
> @@ -1125,6 +1125,9 @@ config ARCH_SELECTS_KEXEC_FILE
>  config ARCH_SUPPORTS_KEXEC_PURGATORY
>  	def_bool ARCH_SUPPORTS_KEXEC_FILE
>  
> +config ARCH_SUPPORTS_KEXEC_HANDOVER
> +	def_bool 64BIT && MMU

[Severity: Medium]
As reported by Sashiko on v1 [1], will kho_reserve_bootmem() fail on 64-bit
RISC-V platforms where DRAM starts at or above 4 GiB, or record a 0-byte
region when all DRAM lies below 4 GiB?

Unlike arm64 (which defines ARCH_LOW_ADDRESS_LIMIT as arm64_dma_phys_limit - 1
and clamps it to memblock_end_of_DRAM() when DRAM starts above 4 GiB), RISC-V
falls back to the default in include/linux/memblock.h:

#ifndef ARCH_LOW_ADDRESS_LIMIT
#define ARCH_LOW_ADDRESS_LIMIT  0xffffffffUL
#endif

However, 64-bit RISC-V supports platforms where all DRAM starts at or above
4 GiB, as seen in arch/riscv/mm/init.c:arch_mm_preinit():

	bool swiotlb = max_pfn > PFN_DOWN(dma32_phys_limit) &&
		       memblock_start_of_DRAM() < dma32_phys_limit;

When KHO is enabled at boot, mm_core_init() -> kho_memory_init() ->
kho_reserve_bootmem() unconditionally allocates both lowmem and global bootmem
without checking whether their computed sizes in bootmem_size_update() are 0:

kernel/liveupdate/kexec_handover.c:kho_reserve_bootmem() {
	...
	size = bootmem_size_lowmem;
	addr = memblock_phys_alloc_range(size, BOOTMEM_ALIGNMENT_BYTES, 0,
					 ARCH_LOW_ADDRESS_LIMIT);
	if (!addr) {
		pr_err("Failed to reserve lowmem bootmem\n");
		goto err_free_bootmem_desc;
	}

	kho_bootmem[i].addr = addr;
	kho_bootmem[i].size = size;
	i++;

	/* reserve large contiguous area for allocations without nid */
	size = bootmem_size_global;
	addr = memblock_phys_alloc(size, BOOTMEM_ALIGNMENT_BYTES);
	if (!addr) {
		pr_err("Failed to reserve global bootmem\n");
		goto err_free_bootmem_areas;
	}
	...
}

On 64-bit RISC-V systems where DRAM starts at or above 4 GiB,
bootmem_size_lowmem is 0 and memblock_phys_alloc_range(0, ..., 0,
ARCH_LOW_ADDRESS_LIMIT) finds no memory below 0xffffffffUL, returning 0 and
disabling KHO. Conversely, when all DRAM lies below 4 GiB, bootmem_size_global
is 0 and kho_reserve_bootmem() records a 0-byte entry in kho_bootmem[], which
later causes kho_locate_mem_hole() to construct an inverted struct resource
with end = start - 1.

[1] https://lore.kernel.org/r/sashiko-outbox-183678@kernel.org

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261103080926.2079-1-cp0613@linux.alibaba.com?part=3