Re: [PATCH v17 03/10] crash: Add crash_prepare_headers() to exclude crash kernel memory

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] crash_prepare_headers() delegates array sizing to architectures but performs memory exclusions in generic code, forcing implicit padding to avoid -ENOMEM.
--

commit 7701a27578ff768dd0a4b481af66ec228b7e613f
Author: Jinjie Ruan <ruanjinjie@huawei.com>

crash: Add crash_prepare_headers() to exclude crash kernel memory

This patch introduces a generic crash_prepare_headers() helper to consolidate
crash memory allocation and exclusions across architectures. It adds
architecture-specific hooks for pre-counting memory ranges, populating the
ranges, and performing any additional exclusions.

> diff --git a/kernel/crash_core.c b/kernel/crash_core.c
> index 4f21fc3b108b8..481babc291310 100644
> --- a/kernel/crash_core.c
> +++ b/kernel/crash_core.c
[ ... ]
> @@ -272,6 +269,85 @@ int crash_prepare_elf64_headers(struct crash_mem *mem, int need_kernel_map,
>  	return 0;
>  }
>  
[ ... ]
> +int crash_prepare_headers(int need_kernel_map, void **addr, unsigned long *sz,
> +			  unsigned long *nr_mem_ranges)
> +{
> +	unsigned int max_nr_ranges;
> +	struct crash_mem *cmem;
> +	int ret;
> +
> +	max_nr_ranges = arch_get_system_nr_ranges();
> +	if (!max_nr_ranges)
> +		return -ENOMEM;
> +
> +	cmem = alloc_cmem(max_nr_ranges);
> +	if (!cmem)
> +		return -ENOMEM;
> +
> +	ret = arch_crash_populate_cmem(cmem);
> +	if (ret)
> +		goto out;
> +
> +	ret = crash_exclude_core_ranges(cmem);

[Severity: Medium]
Will this return -ENOMEM if the generic exclusion splits memory ranges, unless
architectures implicitly pad the count returned by arch_get_system_nr_ranges()?

crash_exclude_core_ranges() calls crash_exclude_mem_range() for crashk_res,
crashk_low_res, and crashk_cma_ranges. If these exclusions split existing
ranges, crash_exclude_mem_range() will attempt to grow cmem->nr_ranges.

Since cmem is allocated exactly to the max_nr_ranges provided by the
architecture, if the architecture doesn't artificially inflate its count to
account for these generic splits, won't the split operations hit the
cmem->nr_ranges >= cmem->max_nr_ranges check and fail with -ENOMEM?

> +	if (ret)
> +		goto out;
> +
> +	ret = arch_crash_exclude_ranges(cmem);
> +	if (ret)
> +		goto out;
> +

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260629094746.191843-1-ruanjinjie@huawei.com?part=3