Re: [PATCH v6 06/10] memblock: introduce MEMBLOCK_NODUMP flag
flat view
From: Baoquan He <baoquan.he@linux.dev>
Date: 2026-09-21 05:53:07
Also in:
kexec, linux-devicetree, linux-iommu, linux-mm, linux-riscv, lkml, loongarch
On 09/20/26 at 04:00pm, Wandun wrote:
On 9/20/26 11:33, Baoquan He wrote:quoted
On 09/19/26 at 10:11am, Mike Rapoport wrote:quoted
On Fri, Sep 18, 2026 at 04:08:10PM +0800, Baoquan He wrote:quoted
On 09/02/26 at 03:31pm, Wandun Chen wrote:...snip...quoted
quoted
quoted
+/** + * memblock_mark_nodump - Mark a memory region with flag MEMBLOCK_NODUMP. + * @base: the base phys addr of the region + * @size: the size of the region + * + * Mark the region as not to be included in crash dumps. + * + * Return: 0 on success, -errno on failure. + */ +int __init_memblock memblock_mark_nodump(phys_addr_t base, phys_addr_t size)~~~~~~~~~~~~~~^ Is it impossible to use at runtime even thought it claims to be generic?This should only happen at boot. Allowing modifications of memblock properties in runtime is a can of worms I don't want to open.Thank you. This is why I said this can't be claimed as a generic way to mark memory region as no dump. The generic way I'd like to see is at any time of system we can mark any unwanted memory region as no dump. And it won't impact the behaviour of the current system.Thanks Mike for the explanation and Baoquan for the suggestion. I'll drop the "generic" related description in the next version. Baoquan, by the way, patch 2 is still awaiting review. It rewrites the riscv side of the code that patch 3 (which you've acked) then folds into the weak defaults, if you could take a look at it too, that would be very helpful.
I checked and looks good, I have added ACK. While RISC-V people can hold it anytime if they have concern.