Received-SPF: pass (google.com: domain of linux-kernel+bounces-132266-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Date: Thu, 4 Apr 2024 16:39:53 -0700
From: Tony Luck <tony.luck@intel.com>
To: Borislav Petkov <bp@alien8.de>, Yazen Ghannam <yazen.ghannam@amd.com>,
	Miaohe Lin <linmiaohe@huawei.com>,
	Naoya Horiguchi <naoya.horiguchi@nec.com>, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org
Subject: Re: Machine check recovery broken in v6.9-rc1
Message-ID: <Zg86SZYDppxffEvv@agluck-desk3>
References: <Zg8kLSl2yAlA3o5D@agluck-desk3>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Zg8kLSl2yAlA3o5D@agluck-desk3>

On Thu, Apr 04, 2024 at 03:05:33PM -0700, Tony Luck wrote:
> v6.8 works. v6.9-rc1 fails with:
> 
> kernel BUG at include/linux/swapops.h:119
> 
> 117 static inline unsigned long swp_offset_pfn(swp_entry_t entry)
> 118 {
> 119         VM_BUG_ON(!is_pfn_swap_entry(entry));
> 120         return swp_offset(entry) & SWP_PFN_MASK;
> 121 }
> 
> 
> I've messed up the bisection three times now and ended up
> pointing at some innocent looking commit. So I wonder if
> anyone else has noticed and tried debugging.

I think I know why I messed up the bisection. I thought
the failure happens every time. But it seems that it only
happens most of the time. I must have accidentally marked
a build as "good" because I got "lucky" and my test passed.

Back to the bisect grindstone :-(

-Tony