Received-SPF: pass (google.com: domain of linux-kernel+bounces-18477-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Message-ID: <f8ae34f0-dab2-48fa-b3f8-5e969ed028b7@linux.intel.com>
Date: Sat, 6 Jan 2024 12:21:49 +0800
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH v3 0/2] iommu/iova: Make the rcache depot properly
 flexible
To: Ido Schimmel <idosch@idosch.org>, Robin Murphy <robin.murphy@arm.com>
Cc: joro@8bytes.org, will@kernel.org, iommu@lists.linux.dev,
 linux-kernel@vger.kernel.org, zhangzekun11@huawei.com,
 john.g.garry@oracle.com, dheerajkumar.srivastava@amd.com, jsnitsel@redhat.com
References: <cover.1694535580.git.robin.murphy@arm.com>
 <ZY1osaGLyT-sdKE8@shredder> <ZZO6NV-Kq-ktnyo5@shredder>
From: Ethan Zhao <haifeng.zhao@linux.intel.com>
In-Reply-To: <ZZO6NV-Kq-ktnyo5@shredder>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit


On 1/2/2024 3:24 PM, Ido Schimmel wrote:
> On Thu, Dec 28, 2023 at 02:23:20PM +0200, Ido Schimmel wrote:
>> On Tue, Sep 12, 2023 at 05:28:04PM +0100, Robin Murphy wrote:
>>> v2: https://lore.kernel.org/linux-iommu/cover.1692641204.git.robin.murphy@arm.com/
>>>
>>> Hi all,
>>>
>>> I hope this is good to go now, just fixed the locking (and threw
>>> lockdep at it to confirm, which of course I should have done to begin
>>> with...) and picked up tags.
>> Hi,
>>
>> After pulling the v6.7 changes we started seeing the following memory
>> leaks [1] of 'struct iova_magazine'. I'm not sure how to reproduce it,
>> which is why I didn't perform bisection. However, looking at the
>> mentioned code paths, they seem to have been changed in v6.7 as part of
>> this patchset. I reverted both patches and didn't see any memory leaks
>> when running a full regression (~10 hours), but I will repeat it to be
>> sure.
> FYI, we didn't see the leaks since reverting these two patches whereas
> before we saw them almost everyday, so I'm quite sure they introduced
> the leaks.

Seems some magazines were not freed when one CPU is dead (hot unplugged) ?

static void free_cpu_cached_iovas(unsigned int cpu, struct iova_domain 
*iovad)
{
     struct iova_cpu_rcache *cpu_rcache;
     struct iova_rcache *rcache;
     unsigned long flags;
     int i;

     for (i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) {
         rcache = &iovad->rcaches[i];
         cpu_rcache = per_cpu_ptr(rcache->cpu_rcaches, cpu);
         spin_lock_irqsave(&cpu_rcache->lock, flags);
         iova_magazine_free_pfns(cpu_rcache->loaded, iovad);

+     iova_magazine_free(cpu_rcache->loaded);

         iova_magazine_free_pfns(cpu_rcache->prev, iovad);

+     iova_magazine_free(cpu_rcache->prev);

         spin_unlock_irqrestore(&cpu_rcache->lock, flags);
     }
}


Thanks,

Ethan

> Thanks
>