Received-SPF: pass (google.com: domain of linux-kernel+bounces-174015-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Received-SPF: Pass (protection.outlook.com: domain of oppo.com designates
 58.252.5.68 as permitted sender) receiver=protection.outlook.com;
 client-ip=58.252.5.68; helo=mail.oppo.com; pr=C
Date: Thu, 9 May 2024 09:30:59 +0800
From: Hailong Liu <hailong.liu@oppo.com>
To: Gao Xiang <hsiangkao@linux.alibaba.com>
CC: <akpm@linux-foundation.org>, Michal Hocko <mhocko@suse.com>,
	<urezki@gmail.com>, <hch@infradead.org>, <lstoakes@gmail.com>,
	<21cnbao@gmail.com>, <linux-mm@kvack.org>, <linux-kernel@vger.kernel.org>,
	<xiang@kernel.org>, <chao@kernel.org>, Oven <liyangouwen1@oppo.com>
Subject: Re: [RFC PATCH] mm/vmalloc: fix vmalloc which may return null if
 called with __GFP_NOFAIL
Message-ID: <20240509013059.vegko7dlty7vppy5@oppo.com>
References: <20240508125808.28882-1-hailong.liu@oppo.com>
 <404add29-2d3f-45db-9103-0c5b66fb254e@linux.alibaba.com>
 <20240508144321.ymnhn54daaabalhe@oppo.com>
 <90a127e5-d884-44b5-bb76-773a7485c2f1@linux.alibaba.com>
 <20240508153136.x4rxildsgza234uv@oppo.com>
 <52f9e300-4ab9-43c1-abae-cbe3da27e5b0@linux.alibaba.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <52f9e300-4ab9-43c1-abae-cbe3da27e5b0@linux.alibaba.com>
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 09 May 2024 01:31:01.0140
 (UTC)
X-MS-Exchange-CrossTenant-Network-Message-Id: 69dac591-7ae9-4d22-69dc-08dc6fc7af5d
X-MS-Exchange-CrossTenant-Id: f1905eb1-c353-41c5-9516-62b4a54b5ee6
X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=f1905eb1-c353-41c5-9516-62b4a54b5ee6;Ip=[58.252.5.68];Helo=[mail.oppo.com]
X-MS-Exchange-CrossTenant-AuthSource:
	SG1PEPF000082E1.apcprd02.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem
X-MS-Exchange-Transport-CrossTenantHeadersStamped: TYZPR02MB5914

On Wed, 08. May 23:40, Gao Xiang wrote:
>
>
> On 2024/5/8 23:31, Hailong Liu wrote:
> > On Wed, 08. May 23:10, Gao Xiang wrote:
> > > Hi,
> > >
> > > On 2024/5/8 22:43, Hailong Liu wrote:
> > > > On Wed, 08. May 21:41, Gao Xiang wrote:
> > > > >
> > > > > +Cc Michal,
> > > > >
> > > > > On 2024/5/8 20:58, hailong.liu@oppo.com wrote:
> > > > > > From: "Hailong.Liu" <hailong.liu@oppo.com>
> > > > > >
> > > > > > Commit a421ef303008 ("mm: allow !GFP_KERNEL allocations for kvmalloc")
> > > > > > includes support for __GFP_NOFAIL, but it presents a conflict with
> > > > > > commit dd544141b9eb ("vmalloc: back off when the current task is
> > > > > > OOM-killed"). A possible scenario is as belows:
> > > > > >
> > > > > > process-a
> > > > > > kvcalloc(n, m, GFP_KERNEL | __GFP_NOFAIL)
> > > > > >        __vmalloc_node_range()
> > > > > > 	__vmalloc_area_node()
> > > > > > 	    vm_area_alloc_pages()
> > > > > >                --> oom-killer send SIGKILL to process-a
> > > > > >                if (fatal_signal_pending(current)) break;
> > > > > > --> return NULL;
> > > > > >
> > > > > > to fix this, do not check fatal_signal_pending() in vm_area_alloc_pages()
> > > > > > if __GFP_NOFAIL set.
> > > > > >
> > > > > > Reported-by: Oven <liyangouwen1@oppo.com>
> > > > > > Signed-off-by: Hailong.Liu <hailong.liu@oppo.com>
> > > > >
> > > > > Why taging this as RFC here?  It seems a corner-case fix of
> > > > > commit a421ef303008
> > > > >
> > > > > Thanks,
> > > > > Gao Xiang
> > > > >
> > > >
> > > > Hi Gao Xiang:
> > > >
> > > > RFC here to wait for a better way to handle this case :).
> > > > IMO, if vmalloc support __GFP_NOFAIL it should not return
> > > > null even system is deadlock on memory.
> > >
> > > The starting point is that kmalloc doesn't support __GFP_NOFAIL
> > > if order > 1 (even for very short temporary uses), see:
> > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/mm/page_alloc.c?h=v6.8#n2896
> > >
> > > but it is possible if we have such page pointer array (since two
> > > (order-1) pages can only keep 1024 8-byte entries, it can happen
> > > if compression ratios are high), and kvmalloc(__GFP_NOFAIL) has
> > > already been supported for almost two years, it will fallback to
> > > order-0 allocation as described in commit e9c3cda4d86e
> > > ("mm, vmalloc: fix high order __GFP_NOFAIL allocations").
> > >
> > > With my limited understanding, I'm not sure why it can cause
> > > deadlock here since it will fallback to order-0 allocation then,
> > > and such allocation is just for short temporary uses again
> > > because kmalloc doesn't support order > 1 short memory
> > > allocation strictly.
> > >
> >
> > deadlock on memory meands there is a memory leak causing
> > system to be unable to allocate memory not actual
> > *deadlock*.
>
> Where is memory leak? If it's caused by kvmalloc(__GFP_NOFAIL)
> callers, then it's bugs of callers and we should fix the callers.
>
> Also why kmalloc(__GFP_NOFAIL) (for example, also order-0
> allocation) differs?
>
> Thanks,
> Gao Xiang
>

I’m not suggesting that erofs would cause a memleak. What I mean is
that if kvmalloc is invoked with __GFP_NOFAIL, it must ensure a non-NULL
return, even in scenarios where memory leaks caused by other processes
result in the inability to allocate a page. In such a situation, it
should result in “Kernel panic - not syncing: System is deadlocked
on memory”.
>
> >
> > > Thanks,
> > > Gao Xiang
> > >
> >
> > --
> >
> > Best Regards,
> > Hailong.

--
Best Regards,
Hailong.