Re: [PATCH v11 1/3] ACPI: APEI: send SIGBUS to current task if synchronous memory error not recovered

From: Shuai Xue <xueshuai@linux.alibaba.com>
To: Borislav Petkov <bp@alien8.de>,
	"james.morse@arm.com" <james.morse@arm.com>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>,
	Dan Williams <dan.j.williams@intel.com>,
	Ira Weiny <ira.weiny@intel.com>,
	"Luck, Tony" <tony.luck@intel.com>,
	rafael@kernel.org, wangkefeng.wang@huawei.com,
	tanxiaofei@huawei.com, mawupeng1@huawei.com,
	linmiaohe@huawei.com, naoya.horiguchi@nec.com,
	gregkh@linuxfoundation.org, will@kernel.org, jarkko@kernel.org,
	linux-acpi@vger.kernel.org, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org, akpm@linux-foundation.org,
	linux-edac@vger.kernel.org, x86@kernel.org, justin.he@arm.com,
	ardb@kernel.org, ying.huang@intel.com, ashish.kalra@amd.com,
	baolin.wang@linux.alibaba.com, tglx@linutronix.de,
	mingo@redhat.com, dave.hansen@linux.intel.com, lenb@kernel.org,
	hpa@zytor.com, robert.moore@intel.com, lvying6@huawei.com,
	xiexiuqi@huawei.com, zhuo.song@linux.alibaba.com
Subject: Re: [PATCH v11 1/3] ACPI: APEI: send SIGBUS to current task if synchronous memory error not recovered
Date: Tue, 27 Feb 2024 09:23:39 +0800	[thread overview]
Message-ID: <e18c242a-6389-4d8c-82f6-30e4606a2cec@linux.alibaba.com> (raw)
In-Reply-To: <20240226102918.GBZdxn_v0NbVjs2XMB@fat_crate.local>

On 2024/2/26 18:29, Borislav Petkov wrote:
> On Sat, Feb 24, 2024 at 02:08:42PM +0800, Shuai Xue wrote:
>> @Borislav, do you have any other concerns?
> 
> Yes, this change needs to be further reviewed by an ARM person: I have
> no clue what those "abnormal synchronous errors" on ARM are 

Hi, Borislav,

May the `abnormal` is not inaccurate and misled you. I mean the preconditions
check before memory_failure_queue():

- `if (!(mem_err->validation_bits & CPER_MEM_VALID_PA))` in ghes_handle_memory_failure()
- `if (flags == -1)` in ghes_handle_memory_failure()
- `if (!IS_ENABLED(CONFIG_ACPI_APEI_MEMORY_FAILURE))` in ghes_do_memory_failure()
- `if (!pfn_valid(pfn) && !arch_is_platform_page(physical_addr)) ` in ghes_do_memory_failure()

If the preconditions are not passed, the user-space process will trigger SEA again.
This loop can potentially exceed the platform firmware threshold or even
trigger a kernel hard lockup, leading to a system reboot.

> and how
> they're supposed to be handled properly there:
> 
> - what happens if you get such an error when ghes is disabled there?

If ghes_disable is set, the GHES driver will not be inited by acpi_ghes_init(),
so none of error notifications will be handled. IMHO, it is expected.

> 
> - is that even the right place to handle them?
> 
> James?
> 

Leave this to @James.

Thank you.

Best Regards,
Shuai