Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Wed, 29 Jan 2020 13:00:16 -0800 (PST)
From:   David Rientjes <rientjes@google.com>
To:     Sandipan Das <sandipan@linux.ibm.com>
cc:     Mina Almasry <almasrymina@google.com>, mike.kravetz@oracle.com,
        shakeelb@google.com, shuah@kernel.org, gthelen@google.com,
        akpm@linux-foundation.org, linux-kernel@vger.kernel.org,
        linux-mm@kvack.org, linux-kselftest@vger.kernel.org,
        cgroups@vger.kernel.org, aneesh.kumar@linux.vnet.ibm.com
Subject: Re: [PATCH v10 7/8] hugetlb_cgroup: Add hugetlb_cgroup reservation
 tests
In-Reply-To: <7ce6d59f-fd73-c529-2ad6-edda9937966d@linux.ibm.com>
Message-ID: <alpine.DEB.2.21.2001291257510.175731@chino.kir.corp.google.com>
References: <20200115012651.228058-1-almasrymina@google.com> <20200115012651.228058-7-almasrymina@google.com> <7ce6d59f-fd73-c529-2ad6-edda9937966d@linux.ibm.com>
User-Agent: Alpine 2.21 (DEB 202 2017-01-01)
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, 23 Jan 2020, Sandipan Das wrote:

> > The tests use both shared and private mapped hugetlb memory, and
> > monitors the hugetlb usage counter as well as the hugetlb reservation
> > counter. They test different configurations such as hugetlb memory usage
> > via hugetlbfs, or MAP_HUGETLB, or shmget/shmat, and with and without
> > MAP_POPULATE.
> > 
> > Also add test for hugetlb reservation reparenting, since this is
> > a subtle issue.
> > 
> > Signed-off-by: Mina Almasry <almasrymina@google.com>
> > 
> 
> For powerpc64, either 16MB/16GB or 2MB/1GB huge pages are supported depending
> on the MMU type (Hash or Radix). I was just running these tests on a powerpc64
> system with Hash MMU and ran into problems because the tests assume that the
> hugepage size is always 2MB. Can you determine the huge page size at runtime?
> 

I assume this is only testing failures of the tools/testing/selftests 
additions that hardcode 2MB paths and not a kernel problem?  In other 
words, you can still boot, reserve, alloc, and free hugetlb pages on ppc 
after this patchset without using the selftests?