Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Message-ID: <bb6473ac-3a12-9667-9425-7b61746e8078@gmail.com>
Date:   Sun, 3 Apr 2022 00:15:04 -0400
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101
 Thunderbird/91.7.0
To:     yaliang.wang@windriver.com
Cc:     akpm@linux-foundation.org, anshuman.khandual@arm.com,
        geert@linux-m68k.org, huangpei@loongson.cn, kumba@gentoo.org,
        linux-kernel@vger.kernel.org, linux-mips@vger.kernel.org,
        penberg@kernel.org, rppt@kernel.org, tsbogend@alpha.franken.de
References: <20220310113116.2068859-1-yaliang.wang@windriver.com>
Subject: Re: [PATCH] MIPS: pgalloc: fix memory leak caused by pgd_free()
Content-Language: en-US
From:   Donald Hoskins <grommish@gmail.com>
In-Reply-To: <20220310113116.2068859-1-yaliang.wang@windriver.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

Hi there,

This fix should backported.  Effectively, all mips64-based Cavium Octeon 
processors have been broken since the original commit; this has been a 
persistent and surreptitious issue for the OpenWrt community since 
nearly a year ago (referencing 
https://forum.openwrt.org/t/upstream-kernel-memleak-5-10-octeon-ethernet-ko/111827/), 
and nearly ended with OpenWrt marking the target as entirely unsupported 
and moving on.

We understand that there are ultimately few users of mips64, but it's 
really important that our memory management work. While OpenWrt can 
backport until upstream inclusion, I am sure there are other users and 
software platforms unaware of this fix (and unable to take advantage of 
it), thus making those platforms on this arch completely unusable.