Re: [RFC Part2 PATCH 07/30] mm: add support to split the large THP based on RMP violation

[Dave Hansen]

On 3/24/21 10:04 AM, Brijesh Singh wrote:
> @@ -1377,6 +1442,22 @@ void do_user_addr_fault(struct pt_regs *regs,
>  	if (hw_error_code & X86_PF_INSTR)
>  		flags |= FAULT_FLAG_INSTRUCTION;
>  
> +	/*
> +	 * If its an RMP violation, see if we can resolve it.
> +	 */
> +	if ((hw_error_code & X86_PF_RMP)) {
> +		ret = handle_rmp_page_fault(hw_error_code, address);
> +		if (ret == RMP_FAULT_PAGE_SPLIT) {
> +			flags |= FAULT_FLAG_PAGE_SPLIT;
> +		} else if (ret == RMP_FAULT_KILL) {
> +			fault |= VM_FAULT_SIGBUS;
> +			mm_fault_error(regs, hw_error_code, address, fault);
> +			return;
> +		} else {
> +			return;
> +		}
> +	}

Won't khugepaged come right back around and coalesce this page again?